NEJM Editorial: Expert says CCTA must endure the natural vetting process

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon

CHICAGO—When the New England Journal of Medicine published the CorE-64 study in the Nov. 26th issue, which produced positive results for cardiac CT angiography, an accompanying editorial questioned whether the technology had enough clinical evidence to warrant widespread CMS reimbursements. Udo Hoffmann, MD, director of CT research at Massachusetts General Hospital in Boston, explains the importance of the trial and analyzes the editorial’s commentary.

The CorE-64 trial—led by Julie Miller, MD, who presented its first results at AHA Scientific Sessions in 2007—found that 64-slice CT scans can correctly identify people with major blood vessel disease and that they are nearly as accurate as invasive coronary angiography. The researchers found that the accuracy of quantitative CT angiography for detecting or ruling out stenoses of 50 percent or more according to conventional angiography revealed an AUC of 0.93, with a sensitivity of 85 percent, a specificity of 90 percent, a positive predictive value of 91 percent and a negative predictive value of 83 percent.

“CorE-64 study is the first and only study that shows better positive predictive value and specificity, than negative predictive value and sensitivity. As a result, the community agrees that strength of CT is through its detection of coronary artery disease,” Hoffmann noted. “CT is currently considered a tool to avoid coronary angiography in patients that are low to intermediate risk.”

However, he concurs with Miller et al’s assessment in the NEJM that CCTA cannot currently be used a tool to replace coronary angiography as a diagnostic tool. “It is really a triage tool for coronary angiography. Once the coronary artery disease (CAD) has been confirmed, many studies are currently being conducted to --identify what the next step should be in quantifying the disease,” Hoffmann said.

“Many studies have revealed that CCTA has a high negative predictive value, so now it just becomes a question of what patient population the technology could be most effective,” he noted.

Hoffmann pointed to the ACCURACY trial, presented last year at RSNA, in which James K. Min et al studied a patient population with a lower prevalence of CAD (about 20-25 percent), compared to CorE-64, which had a disease prevalence of about 56 percent. Miller et al actually studied a patient population in the “higher end of intermediate probability,” according to Hoffmann. Incidentally, the ACCURACY trial found a per-vessel negative predictive value of 64-slice CCTA was 97 percent for identifying blockages greater than 50 percent, and 99 percent for blockages greater than 70 percent.

Because of these variables, Hoffmann suggested that the community does need a large, randomized trial that compares CCTA to current testing methods, such as stress testing.

“While there are guidelines for when to send the patient to receive a stress test, but there are none for CT because of the lack of data,” he pointed out.

In the NEJM editorial, Rita F. Redberg, MD, and Judith Walsh, MD, said that the CorE-64 study “does not advance our knowledge of the appropriate use and possible benefits of the technology.” While Hoffmann said that the study did help to build the data, he noted that this study might not be a reflection of real-life practice, because the various sites involved in the trial, only enrolled about two patients a month, suggesting that the predictive value results might be higher in a busy institution that enrolls many more on a daily basis.

Redberg and Walsh also said that “the continued unrestrained use of new technology, in the absence of evidence-based criteria, portends a bleak future for Medicare and our healthcare system.” They called into question the CMS decision in March to continue widespread coverage of CCTA through its local coverage determination, as opposed to the proposed national coverage determination. The authors suggested that the annual 26 percent increase in cardiac imaging is being fiscally driven, “despite a lack of evidence of outcome benefit.”

Hoffmann disagreed with this commentary about the financial motivation of imaging. “CMS made the right decision based on the expanded data they received. An NCD would have brought this modality to the ground, and would have actually prohibited more research and data, as opposed to encouraging more research,” he said. “CT is going to be a very useful tool, but it takes time and resources to collect a body of literature, from which to make