We evaluated 100 subjects who underwent multislice computed tomography (MSCT) to assess the presence and severity of coronary artery disease (CAD) and to determine the occurrence of coronary events (including cardiac death, nonfatal myocardial infarction, unstable angina requiring hospitalization, and revascularization) over a follow-up period of 16 months.
Assessing a patient's prognosis has become an important part of the treatment of patients with suspected or known coronary artery disease (CAD). Numerous studies have been done on prognostication, employing traditionally used nuclear imaging techniques. In addition, substantial evidence has been gathered on the prognostic value of calcium scoring, supporting the use of this technique for risk stratification in individuals at intermediate risk for CAD.1,2 Contrast-enhanced coronary angiography with multislice computed tomography (MSCT) has recently emerged as an alternative noninvasive modality for the diagnosis of CAD.3 Because the technique permits rapid and noninvasive visualization of the coronary arteries, the procedure is expected to be increasingly used as a first-line evaluation tool.4 Of particular interest is the fact that in addition to the distinction between nonobstructive and obstructive atherosclerosis, the technique can also provide information on plaque composition.5 Whether these observations contain prognostic information, however, is currently unknown. The purpose of our study was to evaluate MSCT angiograms to determine their potential prognostic value.
Subjects and methods
A total of 100 subjects (73 men; average age, 59 years) presenting to the outpatient department and requiring further cardiac assessment were included in the study. MSCT coronary angiography was performed according to the standard protocol,6,7 along with the conventional clinical workup. Exclusion criteria were previous coronary artery bypass graft surgery, contraindications to iodinated contrast, and atrial fibrillation.
Multislice computed tomography angiograms were evaluated on a dedicated workstation for the presence and extent of CAD. The coronary calcium score (based on Agatston8) was determined. The contrast-enhanced coronary angiograms were subsequently evaluated for the presence of CAD. Subjects were considered normal if they did not have any coronary artery calcium or plaques on the contrast scan. Subjects with ≥ 1 coronary plaque on the MSCT scan were considered to have an abnormal scan. Abnormal scans showing at least 1 lesion exceeding 50% luminal narrowing were further categorized as obstructive CAD, and scans showing plaques with < 50% luminal narrowing present were categorized as nonobstructive CAD.
In subjects with obstructive CAD, we documented whether significant lesions were located in the left main (LM) or left anterior descending (LAD) coronary artery. In addition, plaque type was classified as noncalcified, mixed, and calcified. Thus, for each subject, the number of each plaque type, the number of plaques (regardless of stenosis severity), and the number of obstructive plaques were determined. Subjects were followed for the occurrence of cardiac death, nonfatal myocardial infarction (MI), unstable angina requiring hospitalization, and revascularization. Data were analyzed using standard statistical analyses, including Cox regression analysis.
Based on MSCT, 20 subjects did not have CAD. Obstructive CAD was noted in 32 subjects, with involvement of the LM or LAD coronary arteries in 23 subjects. Over a mean follow-up period of 16 months, 26 subjects experienced an event. One subject died of acute MI, nonfatal MI occurred in 3 subjects, unstable angina occurred in 4 subjects, and the remaining subjects underwent revascularization.
Figure 1. First-year event rate
(including death, myocardial infarction,
and unstable angina requiring
hospitalization) for subjects with normal
and abnormal multislice computed
tomography (MSCT) scans.
Figure 2. First-year event rate
(including death, myocardial infarction,
unstable angina requiring hospitalization,
and revascularization) for subjects with
normal and abnormal multislice computed
tomography (MSCT) scans.
In subjects with normal coronary arteries on MSCT, no cardiac events were observed during the first year after MSCT, whereas "hard" events (death, MI, nonfatal MI, and unstable angina) occurred in 5% of subjects with abnormal coronary arteries on MSCT, as shown in Figure 1. When revascularizations were included, the event rate was 30% for subjects with abnormal coronary arteries (Figure 2). When further refining this population by including subjects with either at least 1 significant (50%) stenosis or nonsignificant CAD, event rates were 63% and 8%, respectively. Of note, the event rate was highest (77%) in subjects with significant stenosis in the LM or LAD coronary arteries, as shown in Figure 3.
Subjects who experienced cardiovascular events were significantly older, whereas no differences in risk factors for CAD were noted. On MSCT, subjects presenting with events had a higher calcium score as well as more (obstructive) plaques. Also, relatively more mixed and calcified plaques were present in these subjects compared with event-free subjects. Multivariate analysis (corrected for baseline characteristics with ≤ .5 during univariate analysis) showed that the presence of the following MSCT characteristics were independent predictors of cardiac events: any CAD, obstructive CAD, obstructive CAD in LM or LAD coronary arteries, number of segments with plaques, number of segments with obstructive plaques, and number of segments with mixed plaques.
This study, which provided the first available data on the potential prognostic value of MSCT, showed that observations on MSCT may contain prognostic information. Multivariate analysis adjusted for clinical variables, such as age, sex, and risk factors, showed that this information was incremental to baseline characteristics. The most important finding of the current study, however, was the fact that a normal MSCT was associated with excellent survival.
Previous studies comparing MSCT coronary angiography with conventional coronary angiography have shown that MSCT detects significant (> 50%) stenoses with high diagnostic accuracy.9 In particular, a consistently high negative predictive value (> 90%) has been observed, indicating that the likelihood of CAD is extremely low when MSCT is normal. This notion is further supported by our current prognostic findings. A 100% event-free first-year survival rate was shown in subjects with normal coronary arteries on MSCT. Thus, patients without evidence of atherosclerosis on MSCT may indeed be considered to be at low risk. These patients may then be reassured and discharged without additional testing. Accordingly, the ability to rule out CAD is likely to become the main contribution of MSCT.
Figure 3. First-year event rate in relation to the severity of coronary artery
disease on multislice computed tomography (MSCT) scan. LM indicates left
main coronary artery; LAD, left anterior descending coronary artery.
On the other hand, MSCT variables of plaque burden were shown to be predictive of adverse outcome, incremental to baseline characteristics. Further refinement of these variables showed that event rates rose parallel to increasing extent of disease. Moreover, in line with previous angiographic studies, the highest event rates were seen in subjects with LM or LAD coronary artery stenosis.10 Early detection with MSCT may be of great benefit for these high-risk patients and may allow early initiation of aggressive therapy.
Of additional interest, the presence of nonobstructive stenosis was associated with a slightly but still significantly elevated event rate (compared with the presence of normal coronary arteries). Indeed, plaque rupture has been observed to occur across the full scale of stenosis severity.11-14 Although the risk is highest for more severe lesions, the risk of rupture of nonsignificant lesions is not negligible, as they frequently outnumber the more severe lesions.
Finally, a particularly interesting finding was the association between the number of mixed lesions (containing both noncalcified and calcified tissue) and a worse outcome. Although no studies have linked MSCT plaque type to prognosis thus far, the relation between MSCT plaque type and clinical presentation has been explored in several studies.15,16 These investigations revealed that noncalcified and mixed plaque may be more prevalent in unstable conditions, whereas calcified plaques are typically observed in patients with stable CAD. Potentially, these observations in combination with our current findings may indicate that these mixed plaques may represent more high-risk lesions compared with calcified lesions. Nonetheless, available data are evidently too scarce to support such notions.
This study examined a rather heterogeneous population, and included both subjects with known CAD and those with suspected CAD. As a result, a wide variation in baseline characteristics and treatment strategies was present. More dedicated studies in homogeneous populations are therefore needed. In addition, the present study population was small; validation of our results in larger cohorts is therefore warranted. Finally, MSCT data were visually assessed, and no quantification tools, which could potentially improve reproducibility, are currently available.
Results of our study showed that MSCT coronary angiography has an independent prognostic value over baseline characteristics. The first-year event rate increased with stenosis severity, as well as location in the LM or LAD coronary arteries. However, no events occurred in subjects with a normal MSCT scan, indicating that these subjects have an excellent prognosis.