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

Asymptomatic patients who have an abnormal result on screening coronary computed tomographic angiography (CTA) are more likely to be taking aspirin and statins at 90 days and 18 months than either unscreened patients or those with no signs of atherosclerotic disease upon testing. Positive CTA findings also lead to more secondary testing and invasive revascularization at 90 days without reducing cardiac events at 18 months, according to data published online May 23, 2011, ahead of print in the Archives of Internal Medicine.

The study authors say that the preventive benefit of statins and aspirin is tempered by the risk of further testing in low-risk patients, and thus at this time CTA screening is unjustified.

To investigate the downstream effects of CTA screening, researchers led by Hyuk-Jae Chang, MD, PhD, of Yonsei University Health System (Seoul, South Korea), compared medication use, secondary test referrals, revascularizations, and cardiovascular events in 1,000 asymptomatic patients who underwent CTA from December 2005 to May 2006 and an equal number of matched controls. All subjects were enrolled in the Seoul National University Bundang Hospital health screening program. Those patients who underwent CTA screening were informed of the results at the first visit after the scan.

The mean age of the total study population was 50 years, and 63% were male. The study groups were similar in baseline characteristics, except that controls were more likely to have elevated triglyceride levels (P = 0.004) and lower HDL cholesterol levels (P = 0.02).

Among CTA-screened patients, 79% had a normal result (defined as CTA-negative), while the remaining 21% showed evidence of plaque (CTA-positive). In the latter group, stenosis was significant (≥ 50%) in 5% and severe (≥ 75%) in 2%.

Positive CTA Triggers Increased Statin, Aspirin Use

Prior to enrollment, use of statins and aspirin was similar for the CTA group and controls.

But at the first visit after screening, statins were more commonly prescribed for CTA-positive patients than the control group (P < 0.001), a difference that held across all levels of National Cholesterol Education Program (NCEP) risk for CAD (P = 0.007 for low, P = 0.07 for intermediate, and P = 0.03 for high risk).

The association between positive CTA results and statin prescription translated into increased patient use compared with controls at both 90 days and 18 months (table 1).

Table 1. Statin Use at Baseline and Over Follow-up

On the other hand, those with negative CTA results received fewer statin prescriptions than controls if they had an intermediate NCEP risk (P = 0.03). The CTA-negative group was also less likely than controls to use statins at both 90 days (P = 0.02) and 18 months (P = 0.03).

Multivariate analysis, which included baseline LDL-cholesterol level and history of dyslipidemia, confirmed that statin use was many times greater in the CTA-positive group than controls at 90 days (OR 4.588; 95% CI 2.330-9.036; P < 0.001) and at 18 months (OR 3.307; 95% CI 1.324-8.258; P = 0.01).

A similar pattern was observed for aspirin use. At the index visit, aspirin was more frequently prescribed for CTA-positive patients than controls (P < 0.001). In addition, positive CTA results were linked to increased aspirin use compared with controls over 90-day and 18-month follow-up (table 2).

Table 2. Aspirin Use at Baseline and Over Follow-up

The carryover aspirin use in CTA-positive patients was confirmed by multivariate analysis at 90 days (OR 6.784; 95% CI 3.234-14.231; P < 0.001) and 18 months (OR 4.193; 95% CI 1.825-9.635; P < 0.001). Unlike for statins, there was no trend toward fewer aspirin prescriptions in those with normal CTA results and intermediate NCEP risk compared with controls.

No significant differences were observed in the use of antihypertensive or oral hypoglycemic medications between screened patients and controls over the study period.

More Testing, Revascularization for Screened Patients

At 90 days, referral for secondary tests was more common in the overall CTA group compared with controls (5.5% vs. 2.2%; P < 0.001). Moreover, the difference in referral rates rose along with the NCEP risk (P = 0.60 for low, P = 0.01 for moderate, and P = 0.002 for high risk). However, when the screened group was stratified by imaging results, only 1.4% of CTA-negative patients underwent further testing, whereas 21% of CTA-positive patients were referred for testing (P < 0.01).

Nonetheless, of 55 referrals overall, 11 tests occurred in CTA-negative patients (10 exercise ECGs and 1 angiography) and 22 were performed in controls (13 exercise ECG alone, 4 SPECT alone, 2 angiography alone, and 3 both SPECT and angiography). And despite the increased number of referrals in the entire screened group at 90 days, the percentage of abnormal SPECT tests was no greater than in controls.

In addition, more CTA patients than controls underwent revascularization at 90 days (13 vs. 1 subjects; P < 0.01), although by 18 months the 2 groups had similarly low numbers of interventions (1 CTA patient and 2 controls; P = 0.49). In the first 90 days, there were no cardiac events; after 18 months, 1 patient in the CTA group was admitted for unstable angina while 1 in the control group suffered unspecified cardiac death.

The investigators acknowledge a number of limitations to the study. For one, both subjects and providers were Korean, and their reasons for screening and reaction to CTA results may not be generalizable to other populations. Moreover, the number of cardiovascular events was small, and follow-up was relatively short.

Screening a Double-Edged Sword

Dr. Chang and colleagues observe that the finding of increased test referrals in patients with positive CTA results despite their asymptomatic status and low Framingham Risk Score (placing them in a low 10-year risk group) may reflect a variation of the oculostenotic reflex—a kneejerk reaction to images of stenosis—and raises great concern.

“These findings highlight the need to consider the pretest probability of disease before performing imaging tests in patients who may be subsequently exposed to potentially harmful downstream procedures with questionable prognostic benefit,” they note.

In an invited commentary, Michael S. Lauer, MD, of the National Heart, Lung, and Blood Institute (Bethesda, MD), makes a similar point. The study “serves as a powerful reminder of the 2-edged effects of screening,” he writes, adding that “we do not know whether [coronary CTA] reduced risk of major clinical events, but we now know that it increased exposure to tests and treatments, each of which carries its own risk.”

“The only way to know whether screening by [coronary CTA] leads to clinically beneficial diagnosis of real disease, as opposed to pseudodisease, is by performing large-scale controlled trials, preferably with randomization,” Dr. Lauer advises.

Study coauthor John W. McEvoy, MB, BCh, of Johns Hopkins Ciccarone Center for the Prevention of Heart Disease (Baltimore, MD), underlined in a prepared statement that the issue remains unsettled. “We need longer follow-up because the statin and aspirin therapies have been shown to be beneficial in primary prevention of heart attacks in older patients with multiple cardiac risk factors. In 5 to 10 years, these interventions among the patients who had positive findings on CT angiography may ultimately show some benefit,” he commented. “Right now, we just don’t know.”

In a telephone interview with TCTMD, Michael Poon, MD, of Stony Brook University Medical Center (Stony Brook, NY), said the study highlights the reasons why CTA is not currently indicated for asymptomatic patients. He was surprised that the physicians in the study were so aggressive in this population, noting, “The COURAGE trial suggested that even symptomatic patients should not go to intervention unless they fail optimal medical therapy. So with asymptomatic patients, definitely I would try optimal medical therapy first.”

The paper is especially instructive in that it looks at asymptomatic patients across a spectrum of Framingham risk scores, Dr. Poon said. Even though the cohort included a significant percentage of high-risk patients, the overall rate of stenosis was very low. The lesson is that with CTA “high risk doesn’t necessarily mean high yield,” he observed.

Dr. Poon said the study captures how CTA results can affect patient behavior. While a positive test may motivate some patients to use preventive medications, others may use negative results to justify not taking statins. Although that conclusion is not backed by good data, he said, it does make intuitive sense in that Framingham risk categories are statistical whereas CTA can filter out those who appear on paper to be high risk but in fact are free of CAD.

Currently, use of CTA for risk assessment in asymptomatic patients is not supported by clinical guidelines, Dr. Poon said. And fortunately, he added, after an earlier period when CTA screening was popular, demand for it has waned due to concern about radiation exposure and lack of insurance coverage.

Study Details

In patients undergoing CTA, those with a heart rate higher than 70 beats/min received intravenous esmolol hydrochloride (10-30 mg) before image acquisition. Imaging was performed using a 64-slice multidetector CT scanner (Brilliance 4; Philips Medical Systems, Eindhoven, the Netherlands), using a standard scanning protocol. Scans were analyzed independently on a 3-D workstation by 2 blinded, experienced investigators.

All prescriptions were provided at the discretion of treating physicians. 

2. Lauer MS. Pseudodisease, the next great epidemic in coronary atherosclerosis? Arch Intern Med. 2011;Epub ahead of print.

Related Stories:

• Updated Consensus Statement Outlines Appropriate Cardiac CT Use
• Coronary CTA May Offer Cost-Effective Triage Test in Asymptomatic Patients
• CT Angiography Rules Out Stenosis in Patients at Low-Intermediate CAD Risk