ASNC: CT Angiography Improves Accuracy of Myocardial Perfusion Imaging

SAN DIEGO, Sept. 14 -- For patients with equivocal or discordant findings on myocardial perfusion imaging, CT angiography (CTA) can unmask false-negative and false-positive results, investigators reported here.

CTA improved the specificity and positive predictive value of perfusion imaging results that were equivocal or abnormal, Regina S. Druz, M.D., of North Shore University Hospital in Manhasset, N.Y., told attendees at the American Society of Nuclear Cardiology meeting.

The study suggested two patient populations in whom CTA might be especially beneficial in clarifying perfusion imaging results.

"In patients with normal or mildly abnormal results, CTA demonstrated moderate or severe coronary artery disease not identified by [myocardial perfusion imaging parameters], thus identifying false-negative patients," said Dr. Druz. "CTA identified patients who had no or mild stenosis, thus identifying false-positive [perfusion imaging] results."

Myocardial perfusion imaging has an established role in the diagnosis and risk stratification of coronary artery disease, Dr. Druz said. Multidetector CTA has demonstrated accuracy in characterizing coronary anatomy compared with coronary angiography. But, she noted, whether CTA can improve the accuracy of MPI has not been studied extensively.

So Dr. Druz and colleagues retrospectively reviewed records of 74 patients who underwent both perfusion imaging and CTA. The cohort included 22 patients with known coronary artery disease, 14 of whom had prior percutaneous coronary intervention.

CTA and perfusion imaging were performed within a year of each other (median of 29 days). Perfusion imaging preceded CTA in 58 cases, and the 22 patients with coronary artery disease all had conventional coronary angiography. Dr. Druz reported that 51 patients had exercise stress tests, and the remaining 23 had pharmacologic stress tests.

Perfusion imaging results were interpreted on the basis of summed stress scores (SSS) and summed difference scores (SDS) derived from a five-point/20-segment model. A segmental score of 0 reflected normal tracer uptake, and a score of 4 indicated absence of uptake.

The calculations led to categorization of myocardial perfusion imaging results in the following manner:

• Normal: SSS = 0-1
• Mildly abnormal: SSS = 2-3 (rest perfusion normal); SSS = 4-8 and SDS = 0-1 (rest perfusion abnormal)
• Markedly abnormal: SSS 9+ (rest perfusion normal); SSS = 4-8 and SDS 2+ (rest perfusion abnormal)

CTA was performed with a 64-slice machine, using standard protocols for contrast-enhanced and noncontrast-enhanced scans. Among the patients who had coronary angiography, a stenosis of 70% or greater in any of the major arteries or side branches was considered significant, said Dr. Druz.

The patients included in the study had a median age of 62.5, and 56 of the 74 had been referred for imaging because of chest pain and/or dyspnea. The remaining patients consisted of one referred for preoperative evaluation, one with known coronary disease, and six referred for perfusion imaging because of abnormal CT results.

Perfusion imaging results were normal in 18 cases, mildly abnormal in 24, and markedly abnormal in 32. CTA in those patients revealed stenosis in seven of 18 with normal perfusion imaging (moderate or severe in five of seven); 14 of 24 with mildly abnormal imaging results (moderate-severe in nine); and 21 of 32 with abnormal perfusion imaging (moderate-severe in 15).

Dr. Druz noted that the likelihood of finding stenosis with CTA was significantly greater in patients with abnormal myocardial perfusion imaging compared with normal or mildly abnormal imaging results (P=0.03).

Of the 22 patients with known coronary artery disease, angiography revealed stenosis of 70% or greater in 13. CTA correctly identified 14 of the 22 patients as having disease, and perfusion imaging identified 13.

For patients with coronary stenosis of 50% or greater (moderate-severe) by angiography, CTA added to equivocal or mildly abnormal imaging boosted specificity from 25% with perfusion imaging alone to 75% and positive predictive value from 70% to 82% (P=0.004 for both comparisons).

Among patients with frankly abnormal myocardial perfusion imaging results, CTA increased specificity from 63% to 100% and positive predictive value from 70% to 100%, but the differences were not statistically significant because of the small number of patients.

"The severity of coronary artery disease as defined by CTA correlated with perfusion imaging findings," said Dr. Druz. "With increasing [SSS/SDS], we observed a greater prevalence of moderate or severe coronary stenosis on CTA."