Prognostic Value of Coronary Magnetic Resonance Angiography for Prediction of Cardiac Events in Patients With Suspected Coronary Artery Disease
Author + information
- Yeonyee E. Yoon, MD⁎,†,
- Kakuya Kitagawa, MD⁎,
- Shingo Kato, MD⁎,
- Masaki Ishida, MD⁎,
- Hiroshi Nakajima, MD‡,
- Tairo Kurita, MD§,
- Masaaki Ito, MD‡ and
- Hajime Sakuma, MD⁎,⁎ (sakuma{at}clin.medic.mie-u.ac.jp)
- ↵⁎Reprint requests and correspondence:
Dr. Hajime Sakuma, Department of Radiology, Mie University Hospital, 2-174 Edobashi, Tsu, Mie 514-8507, Japan
Abstract
Objectives This study sought to determine whether whole-heart coronary magnetic resonance angiography (CMRA) can predict cardiac events in patients with suspected coronary artery disease.
Background Recent studies demonstrated that the presence of stenosis on coronary computed tomography angiography has a significant prognostic impact on the prediction of cardiac events. However, the prognostic value of whole-heart CMRA is unknown.
Methods We studied 207 patients with suspected coronary artery disease who underwent non-contrast-enhanced free-breathing whole-heart CMRA acquired with a 1.5-T MR system and 32-channel cardiac coils. The presence of significant coronary stenosis (≥50% diameter reduction) was visually determined on sliding thin- maximum intensity projection images. Follow-up information was obtained for occurrence of severe cardiac events (cardiac death, myocardial infarction, and unstable angina) and all cardiac events (additionally including revascularization>90 days after CMRA).
Results During a median follow-up of 25 months, 10 cardiac events, of which 5 were severe, were observed in 84 patients with significant stenosis. Whereas, in 123 patients without significant stenosis, only 1 cardiac event with no severe event was observed. Kaplan-Meier curves demonstrated a significant difference in event-free survival between the 2 groups for severe events (annual event rate, 3.9% and 0%, respectively; log-rank test, p = 0.003), as well as for all cardiac events (6.3% and 0.3%; p < 0.001). Cox regression analysis showed that presence of significant stenosis on CMRA was associated with a >20-fold hazard increase for all cardiac events (hazard ratio: 20.78; 95% confidence interval: 2.65 to 162.70; p = 0.001).
Conclusions Whole-heart CMRA is useful for predicting the future risk for cardiac events in patients with suspected coronary artery disease.
Coronary artery disease (CAD) remains the leading cause of morbidity and mortality in developed countries (1). The availability of effective treatment options by both medical and interventional therapy makes it important to identify patients at risk, and it is equally important to identify patients who do not need treatment. Although conventional coronary X-ray angiography is the current gold standard for the detection of CAD, it is expensive, potentially harmful, and associated with a small risk of serious complications. Coronary computed tomography (CT) angiography offers noninvasive approach for evaluation of coronary arteries and showed high diagnostic accuracy to detect and exclude obstructive CAD (2–4). In addition, recent studies demonstrated that the presence of stenosis on coronary CT angiography is associated with increased adverse cardiac events in patients with suspected CAD (5–8). However, coronary CT angiography has limitations in terms of use of iodinated contrast agent and radiation exposure.
Coronary magnetic resonance angiography (CMRA) has emerged as a possible noninvasive alternative for visualizing coronary arteries (9). Recently, 3-dimensional (3D) steady-state free precession (SSFP) whole-heart CMRA, which allows visualization of all major coronary arteries with a single axial 3D acquisition, has become the method of choice for MR coronary imaging (10–13). Because of an intrinsically high blood signal intensity (14), SSFP whole-heart CMRA images can be acquired without administering contrast medium. Several studies evaluated the diagnostic performance of 1.5-T SSFP whole-heart CMRA for detecting significant coronary artery stenosis on invasive coronary angiography (15–18). However, to date, data on the prognostic value of whole-heart CMRA are not available. Therefore, in this study, we investigated whether CMRA can predict cardiac events in patients with suspected CAD.
Methods
Study population
The medical records were reviewed in 278 consecutive adult patients with suspected CAD who underwent non-contrast-enhanced free-breathing whole-heart CMRA between January 1, 2007 and December 31, 2008. The acquisition and analysis of CMRA was approved by the Institutional Review Board and written informed consent was obtained from all subjects at the time of CMRA. Of the 278 patients who had CMRA, 52 patients with known prior CAD at the time of CMRA, including prior myocardial infarction, CAD on prior catheterization, or prior revascularization were excluded. Thus, the presence of significant stenosis on CMRA was assessed and the clinical information was obtained in the remaining 226 patients with suspected CAD but without previously known CAD. The Institutional Review Board had approved this retrospective study to determine the prognostic value of CMRA and waived the need for additional written informed consent.
Image acquisition
MR images were obtained by using a 1.5-T MR imaging unit (Achieva, Philips Medical Systems, Best, the Netherlands) with 32-channel cardiac coils. Isosorbide dinitrate (5 mg) was administered sublingually to all subjects before CMRA acquisition. Beta-blockers were not used in this study. Initial survey images were obtained to determine the position of the heart and diaphragm. Then, reference images were acquired to evaluate the individual coil sensitivities for parallel imaging acquisition while the patient breathed freely. To monitor motion of the right coronary artery, transaxial cine MR images were acquired using SSFP sequence while breathing freely (repetition time: 2.6 ms, echo time: 1.3 ms, flip angle: 60°, field of view: 320 × 320 × 120 mm, acquisition matrix: 128 × 128, cardiac phases: 50, sensitivity encoding factor: 3.0, imaging time: 3 s). A patient-specific acquisition window was set during either systole or diastole, depending on the phase of minimal motion of the right coronary artery (13,15). Free-breathing, navigator-gated whole-heart CMRA was obtained using a 3D, segmented SFFP sequence with radial k-space sampling (repetition time: 4.6 ms, echo time: 2.3 ms, flip angle: 90°, full Fourier encoding, excitations per cardiac cycle: 20 to 50, navigator gating window: ±2.5 mm, field of view: 280 × 280 × 120 mm, acquisition matrices: 256 × 256 × 80, reconstruction matrices: 512 × 512 × 160). Myocardial and venous blood signal were suppressed using T2 preparation. Spectral pre-saturation with inversion recovery was also applied to suppress epicardial fat signals.
Image analysis
CMRA images were transferred to a 3D image server (Aquarius NET server, Tera-Recon, Inc., San Mateo, California). The coronary artery tree was segmented according to the modified American College of Cardiology/American Heart Association classification (19).
Two observers, blinded to clinical information, evaluated the entire coronary arteries on whole-heart CMRA by using sliding thin-slab maximum intensity projection to determine the presence or absence of significant luminal narrowing (≥50% diameter reduction) (15,18,20). Disagreement between the 2 observers was settled by a consensus reading. All coronary arteries were included for the evaluation regardless of the image quality of CMRA, by using an intention-to-read approach.
Follow-up
Follow-up information was obtained by either clinical visits or telephone interviews. All reported events were verified by hospital records or direct contacts with the attending physicians. The following clinical events were recorded: 1) cardiac death; 2) nonfatal myocardial infarction; 3) unstable angina pectoris requiring hospitalization; and 4) coronary revascularization (by either bypass surgery or percutaneous coronary intervention). Cardiac death was defined as death caused by acute myocardial infarction, ventricular arrhythmias, or refractory heart failure. In addition, information on noncardiac deaths was also collected. Coronary revascularizations occurring early after CMRA were usually performed as a result of CMRA because patients with significant coronary artery stenosis were recommended to have invasive angiography and intervention as needed (7,21). Therefore, coronary revascularization procedures performed within 90 days after CMRA were reported separately as an outcome of clinical interest and were excluded from cardiac events. For follow-up analysis, all patients undergoing early revascularization were censored at the time of the intervention and were therefore excluded from further analysis.
Statistical analysis
Continuous variables are expressed as mean ± SD and compared using Student t test. Age was the only continuous variable and was normally distributed. Categorical variables are expressed as proportions and compared by the chi-square test or Fisher exact test, as appropriate. A 2-sided p value of <0.05 was considered statistically significant. Kaplan-Meier curves were used to estimate the distribution of time to cardiac events according to the presence or absence of significant stenosis. Differences between time-to-event curves were compared with the log-rank test. Annual event rates were calculated by dividing the 3-year Kaplan-Meier event rates by 3. The effect of the significant stenosis diagnosed by CMRA on all cardiac events was determined using the Cox regression analysis. Results are reported as hazard ratios (HR) with 95% confidence intervals (CI). All analyses were performed with the SPSS statistical package (version 19.0, SPSS, Inc., Chicago, Illinois).
Results
Of the 226 patients, 19 (8.4%) patients who were lost to follow-up were excluded. Thus, 207 patients (128 men; mean age: 68 ± 10 years) were available for the final analysis. All 3 major coronary arteries were visualized in all study subjects. The number of segments that were visible and evaluated in this study, using a modified American College of Cardiology/American Heart Association classification (19), was described in Online Table 1. The most common indication for CMRA was chest pain (101 [49%]), and other indications included dyspnea (14 [7%]), palpitation (18 [9%]), and syncope (6 [3%]) (Figs. 1 and 2).⇓⇓ Asymptomatic patients (68 [33%]) mainly consisted of those with multiple risk factors. The detailed patient characteristics and indication for CMRA are shown in Table 1. Whole-heart CMRA revealed at least 1 significant coronary artery stenosis in 84 of 207 studied patients (41%). Of the 84 patients with significant stenosis, 47 patients had significant CAD in only 1 coronary artery, 25 in 2 arteries, and 12 had significant 3-vessel disease. Patients with significant stenosis on CMRA were older and more likely to be men (Table 1).
Images in a 81-Year-Old Male Patient With Multiple Risk Factors
(A) Thin-slab maximum intensity projection image and (B) conventional angiogram show significant stenosis (red arrow) in the proximal portion of the left anterior descending artery.
Images in a 84-Year-Old Male Patient With Chest Pain on Effort
(A) Thin-slab maximum intensity projection image and (B) conventional angiogram show significant stenosis (red arrow) in the proximal portion of the left circumflex artery.
Patient Characteristics
The median follow-up period was 25 months (interquartile range 16 to 33 months). The cardiac events during follow-up were summarized in Table 2. A total of 5 severe cardiac events (cardiac death: n = 1; unstable angina: n = 4) were observed in patients with significant stenosis, whereas none of the patients without significant stenosis on CMRA experienced any severe cardiac events (log-rank p = 0.003). Annualized rate of severe cardiac events was 3.9% and 0% for patients with and without significant stenosis, respectively. Kaplan-Meier curves, stratified by the presence of significant stenosis for severe cardiac events are illustrated in Figure 3A. Noncardiac death occurred in 1 patient without significant stenosis (malignancy: n = 1) and in 2 patients with significant stenosis (cerebrovascular event: n = 1; trauma: n = 1) (p = 0.232). Those with noncardiac death were censored at the time of death.
Kaplan-Meier Event-Free Survival Curves
Curves for (A) severe cardiac events and (B) all cardiac events.
Follow-Up Events and Combined Endpoints
A total of 25 patients with significant stenosis (25 of 84 [29.8%]) underwent revascularization (percutaneous coronary intervention: n = 19, bypass surgery: n = 6) within 90 days after CMRA. In contrast, no one without significant stenosis underwent early revascularization within 90 days after CMRA. All patients undergoing early revascularization were censored at the time of intervention for the survival analysis. During follow-up, 8 patients underwent late (>90 days after CMRA) revascularization (percutaneous coronary intervention: n = 7, bypass surgery: n = 1). Significantly more late coronary revascularization procedures were performed in patients with significant stenosis than in those without significant stenosis (7 vs. 1 revascularization procedure in patients with significant stenosis vs. no significant stenosis, respectively; p = 0.001). All cardiac events, including cardiac death, myocardial infarction, unstable angina, and late revascularization were observed in 10 of 84 (11.9%) patients with significant stenosis and only in 1 of 123 (0.8%) patients without significant stenosis, corresponding to the annual event rate of 6.3% and 0.3%. Kaplan-Meier curves showed worse event-free survival in patients with significant stenosis than in those without significant stenosis (p < 0.001) (Fig. 3B).
Univariate analyses of clinical variables and the presence of significant stenosis for composite endpoints of all cardiac events are summarized in Table 3. As no severe cardiac event was observed in patients without significant stenosis, Cox regression analysis was available only for all cardiac events. Although current smoking (unadjusted HR: 4.88, 95% CI: 1.43 to 16.67; p = 0.011) and family history of premature CAD (unadjusted HR: 3.95; 95% CI: 1.04 to 14.99; p = 0.044) were significant predictors of all cardiac events, the presence of significant stenosis was the strongest predictor of all cardiac events (unadjusted HR: 20.78, 95% CI: 2.46 to 145.40; p = 0.004). When adjusted to current smoking and family history of premature CAD, presence of significant stenosis maintained a >17-fold adjusted hazard increase for all cardiac events (adjusted HR: 17.97, 95% CI: 2.28 to 141.74; p = 0.006).
Univariate Analysis of Factors Associated With All Cardiac Events
Discussion
The present study demonstrated the following findings: 1) the frequencies of severe cardiac events as well as all cardiac events were significantly higher in patients with significant stenosis on CMRA in comparison with patients without significant stenosis; 2) the absence of significant stenosis on CMRA can identify a population with a very low risk for severe cardiac events (0%) and all cardiac events (0.3%) during a median follow-up period of 25 months.
At present, coronary CT angiography appears to be the most robust technique for noninvasive visualization of the coronary arteries. Sixty-four–slice multidetector CT has been shown to be a sensitive and specific tool for the detection of significant coronary stenosis and has been validated against conventional coronary angiography (2–4). Although use of CMRA for assessing CAD is currently limited due to lower spatial resolution (1 to 1.5 mm) and lengthy imaging time (typically 5 to 15 min) and complexity of examination that can lead to inconsistent image quality (15,18,20), CMRA has several advantages over coronary CT angiography (22). First, CMRA does not expose the patients to ionizing radiation. Second, no contrast material injection is required to obtain 1.5-T SSFP CMRA. Third, the lumen of the coronary artery can be visualized in patients with heavy coronary artery calcification. The diagnostic accuracy of whole-heart CMRA has been investigated in several single-center studies, with a sensitivity of 78% to 96% and a specificity of 68% to 96% in the detection of coronary arteries with a luminal narrowing of at least 50% identified at conventional angiography (15–17). In a multicenter trial by Kato et al. (18), 1.5-T whole-heart CMRA acquired with 5-channel cardiac coils showed a high negative predictive value of 88%, indicating that CMRA is useful for ruling out CAD in patients with suspected CAD and thus is of value for eliminating unnecessary X-ray angiography. In addition, 1.5-T whole-heart CMRA reliably ruled out left main disease or 3-vessel disease with a high negative predictive value of 99%. Although prolonged imaging time has been a major disadvantage of whole-heart CMRA, introduction of 32-channel cardiac coils substantially reduced imaging time of CMRA, with the vessel-based sensitivity of 86% and specificity of 93% (20).
An understanding of the prognostic value as well as the diagnostic accuracy of the new imaging technique is important before adopting it into clinical practice. During the last few years, several studies reported prognostic value of coronary CT angiography (5–7,23,24). In a study by Hadamitzky et al. (7), the frequencies of severe and all cardiac events were significantly higher in patients with significant coronary stenosis on coronary CT angiography (event rate in the first year: 1.7% and 4.3%, respectively) in comparison with patients without significant stenosis (0.1% and 0.3%, respectively). To our knowledge, the current study is the first to evaluate the prognostic value of whole-heart CMRA. Patients with significant stenosis by CMRA suffer significantly more frequently from severe and all cardiac events (annual event rate: 3.9% and 6.3%, respectively) than do patients without significant stenosis (0% and 0.3%, respectively). Furthermore, the high hazard ratio of 20.78 observed in this study indicated that the presence of significant stenosis on whole-heart CMRA is strongly associated with future major cardiac events in patients with suspected CAD.
Some studies showed that many acute myocardial infarctions are caused by lesions classified as nonobstructive in previous angiograms (25,26). However, the current whole-heart CMRA approach does not provide sufficient spatial resolution to stratify the lesions into mild, moderate, and severe stenosis. Methods that allow for more quantitative analysis of luminal narrowing need to be established in the future studies. Nevertheless, in the current study, the presence of lumen narrowing by the visual assessment of CMRA is proven useful for the prediction of severe cardiac events including cardiac death, acute myocardial infarction, and unstable angina.
It is of great clinical importance that the absence of significant stenosis on CMRA can safely rule out the future risk for cardiac events in symptomatic individuals. Recent studies consistently showed annual death or myocardial infarction rates <0.5% in patients without significant CAD on coronary CT angiography (5–8). In the present study, we found that the absence of significant stenosis on CMRA was also associated with excellent prognosis with an annualized rate of 0.3% for all cardiac events. This finding is of major clinical relevance, because these patients may indeed be safely reassured without the need for further testing.
Study limitations
Our single-center study is limited by the relatively small number of patients affecting the precision and statistical power of our analysis and therefore needs to be confirmed in a larger patient population. A small number of adverse cardiac events preclude a valid multivariate analysis for comparison of CMRA with other risk predictors. In addition, the present study was performed in patients with intermediate CAD prevalence of 41%. The findings in this study may not be directly extrapolated to the population with a lower prevalence of CAD. Further studies are required to examine the prognostic value of CMRA in asymptomatic individuals. Finally, this study was performed as a retrospective study. The diagnostic and therapeutic procedures were not guided by a specific study protocol and might be influenced by a result of CMRA. However, such effects are inevitable in a study observing clinical treatment pathways. And, in the present study, the difference in the event rates was not influenced by complications of revascularization, because patients were censored for the endpoint of all cardiac events after the first revascularization.
Conclusions
Our current study demonstrates the capacity of CMRA for the prediction of cardiac events in patients with suspected CAD. The high hazard ratio observed in this study indicated that the presence of significant stenosis on whole-heart CMRA is strongly associated with future major cardiac events in patients with suspected CAD. In addition, the absence of significant stenosis on whole-heart CMRA is associated with a very low risk of cardiac events. Whole-heart CMRA permits the risk stratification of patients with suspected CAD without exposing the subjects to radiation or administrating contrast medium.
Appendix
For a supplemental table, please see the online version of this paper.
Appendix
Prognostic Value of Coronary Magnetic Resonance Angiography for Prediction of Cardiac Events in Patients With Suspected Coronary Artery Disease
[S0735109712045020_mmc1.docx]Footnotes
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- 3D
- 3-dimensional
- CAD
- coronary artery disease
- CI
- confidence interval(s)
- CT
- computed tomography
- CMRA
- coronary magnetic resonance angiography
- HR
- hazard ratio(s)
- SSFP
- steady-state free precession
- Received February 3, 2012.
- Revision received June 14, 2012.
- Accepted July 3, 2012.
- American College of Cardiology Foundation
References
Figures in Article
Figures
Supplementary Materials
- Online Table 1
Prognostic Value of Coronary Magnetic Resonance Angiography for Prediction of Cardiac Events in Patients With Suspected Coronary Artery Disease
[S0735109712045020_mmc1.docx]