Cardiac Imaging • Original Rese arch

Ghersin et al. MDCT Versus Invasive Coronary Angiography in Chest Pain

A

C E N T U R Y

MEDICAL

O F

IMAGING

Eduard Ghersin1 Diana Litmanovich1 Robert Dragu2 Shmuel Rispler2 Jonathan Lessick2 Amos Ofer1 Olga R. Brook1 Luis Gruberg2 Rafael Beyar2 Ahuva Engel1

Keywords: angiography, chest pain, coronary artery disease, MDCT angiography DOI:10.2214/AJR.04.1232 Received August 3, 2004; accepted after revision December 21, 2004. 1Department

of Diagnostic Imaging, Rambam Medical Center, Haifa, Israel. Address correspondence to E. Ghersin ([email protected]).

2Department

of Cardiology, Rambam Medical Center,

Haifa, Israel. AJR 2006; 186:177–184 0361–803X/06/1861–177 © American Roentgen Ray Society

AJR:186, January 2006

16-MDCT Coronary Angiography Versus Invasive Coronary Angiography in Acute Chest Pain Syndrome: A Blinded Prospective Study OBJECTIVE. The purpose of our study was to prospectively evaluate the usefulness of CT coronary angiography versus invasive coronary angiography for the detection of clinically significant coronary artery disease in patients hospitalized for acute chest pain syndrome. SUBJECTS AND METHODS. Sixty-six consecutive patients (52 men and 14 women; average age, 57 ± 11 [SD] years) who were hospitalized for acute chest pain syndrome underwent CT coronary angiography and invasive coronary angiography within an average time interval of 4 days. ECG-gated CT coronary angiography was performed with a 16-MDCT scanner (0.42-sec rotation time, 16 × 0.75 mm detector collimation). Beta-blockers were not administered routinely, and thus the average heart rate was 71 ± 11 beats per minute. CT coronary angiographic images were evaluated concurrently by two radiologists, who were blinded to invasive coronary angiography results, for stenoses having a diameter of 50% or more, using a 15-segment classification, including all segments 2 mm or more in diameter. The consensus interpretation was compared with results of invasive coronary angiography. RESULTS. CT coronary angiography was technically successful in 59 patients (89%). After exclusion of 20 (3.1%) of 649 coronary segments, which were classified as nonevaluable by CT coronary angiography, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT coronary angiography for identifying significant coronary artery disease in the remaining 629 coronary segments were 80% (68/85), 89% (482/544), 52% (68/130), 97% (482/499), and 87% (550/629), respectively. The overall accuracy for the main vessels (left main, left anterior descending, left circumflex, and right coronary arteries) was 93%, 88%, 86%, and 86%, respectively. CONCLUSION. CT coronary angiography using a 16-MDCT scanner enables accurate noninvasive detection of significant coronary artery disease in patients hospitalized for acute chest pain syndrome. Furthermore, relative high sensitivity and specificity of CT coronary angiography can be achieved without pharmacologic manipulation of patient heart rates. ince the advent of MDCT scanners with their improved spatial, contrast, and temporal resolutions, CT coronary angiography has been gradually evolving as a promising noninvasive method for the assessment of patients with possible coronary artery disease [1–4]. Nevertheless, performance of CT coronary angiography versus that of invasive coronary angiography, especially in patients with acute chest pain syndrome, still needs to be validated. For this study, our primary objective was to evaluate the sensitivity, specificity, positive and negative predictive values, and accuracy of CT coronary angiography in identifying significant coronary artery disease (CAD) using invasive coronary angiography as a gold standard, in a prospective study of patients hospi-

S

talized for acute chest pain syndrome. Because all previously published studies evaluated CT coronary angiography in patients with stable angina pectoris [5], we wanted to test the role of CT coronary angiography in the clinical setting of acute chest pain syndrome. In addition, we defined three secondary objectives. The first was to determine whether high accuracy could be achieved when performing a study using non-pharmacologically-modulated heart rates. This objective was selected because all previously published studies evaluated CT coronary angiography in conjunction with the administration of β–blockers, resulting in relative reduced heart rates [3–5]. The second secondary objective was to determine whether high accuracy could be achieved when performing CT coro-

177

Ghersin et al. nary angiography in patients with a substantial arterial calcium burden as reflected by the individual arterial calcium Agatston score. This objective was selected because arterial calcium is a known major pitfall in the evaluation of coronary artery stenoses, especially using 4-MDCT scanners. It has been postulated that the improved spatial resolution offered by 16-MDCT scanners may overcome this limitation [6]. The third secondary objective was to evaluate the sensitivity and accuracy of CT coronary angiography in identifying coronary stent patency and significant instent luminal stenosis. This objective was selected because previous studies using 4MDCT scanners found reliable evaluation of in-stent stenoses to be impractical [7]. Subjects and Methods Study Population Between February 2003 and March 2004, a group of 66 consecutive patients who were hospitalized for acute chest pain syndrome underwent both CT coronary angiography and invasive coronary angiography as close as clinically possible. Except for seven patients, both CT coronary angiography and invasive coronary angiography were performed within an interval of 7 days. The overall average time between the two examinations was 4 days. Thirty-three patients (50%) underwent CT coronary angiography before invasive coronary angiography, and 33 patients (50%) underwent CT coronary angiography after invasive coronary angiography. The study group included 52 men and 14 women who were 34–79 years old (average age, 57 ± 11 [SD] years). Most enrolled patients were ultimately diagnosed as having acute coronary syndrome; 40 (60.6%) were diagnosed as having acute myocardial infarction and 19 (28.8%) as having unstable angina pectoris. Only seven patients (10.6%) were ultimately diagnosed as having nonspecific chest pain. The only criteria for exclusion were allergy to iodine contrast media, renal insufficiency (serum creatinine > 1.4 mg/dL), and arrhythmia. Prior cardiac surgery, coronary arterial interventions, and Agatston calcium scores did not serve as exclusion criteria.

We continuously injected 120–130 mL of nonionic iodinated contrast material (Ultravist [iopromide], 370 mg I/mL, Schering) at a rate of 4 mL/sec using a power injector. ECG-gated calcium scoring was added to the imaging protocol during the second half of the study, in 29 patients, to detect and quantify the amount of calcification in the coronary arteries. The following imaging and reconstruction parameters were used: collimation, 8 × 3 mm; kVp, 120; mAs, 55; pitch, 0.2; rotation time, 0.42 sec; slice width, 3 mm; and increment, 24 mm. The postprocessing reformations and measurements were performed on a Brilliance Extended Workspace workstation (Philips Medical Systems). We calculated the total Agatston calcium score. A team of two radiologists, who were blinded to the patient’s history and the invasive coronary angiography results, evaluated the CT coronary angiography in consensus using a modified American College of Cardiology/American Heart Association (ACC/AHA) classification, including all segments 2 mm or greater in diameter belonging to the left main, left anterior descending, left circumflex, and right coronary arteries [8]. CT coronary angiographic images were reformatted using curved multiplanar reformations through the lumen of the coronary vessels. On the basis of established cardiology practice, the minimal diameter of stenotic coronary lesions was measured using manually positioned electronic calipers and compared with the maximal diameter of the closest proximal normal arterial segment. Segments were graded as normal, less than 50% stenosis, 50–70% stenosis, or greater than 70% stenosis. According to accepted cardiology practice, stenoses having a diameter equal to or greater than 50% were regarded as significant and these results were compared with results of invasive coronary angiography [9]. Coronary stents (n = 27) were evaluated using curved multiplanar reformations through the stent lumen. For optimal visualization of the stent lumen, we used a bone window setting of approximately 300-H window level and 1,000-H window width. Additional minor window adjustments for each individual stent were made to optimize the distinction between the stent strut and stent lumen. Coronary stents were classified into one of three categories: patent, occluded, or having significant in-stent luminal narrowing.

CT Protocol and Image Reconstruction Unlike prior studies, routine premedication with β–blockers for lowering heart rate was not used. ECG-gated CT coronary angiography studies were performed with a Brilliance 16-MDCT scanner (Philips Medical Systems). The following imaging and reconstruction parameters were used: detector collimation, 16 × 0.75 mm; kVp, 140; mAs, 400–500; pitch, 0.2; rotation time, 0.42 sec; slice width, 1 mm; and increment, 0.5 mm.

178

Invasive Coronary Angiography Protocol Conventional invasive coronary angiography was performed using standard techniques. Images were acquired in optimal projection angles, at 25 frames per second, and were digitally recorded on a Coroskop Top system (Siemens Medical Solutions). Experienced interventional cardiologists, who were blinded to CT coronary angiography results and to patients’ histories, interpreted all angio-

grams using manually positioned electronic calipers. Each coronary angiogram was screened for lesions with stenoses equal to or larger then 50% of the lumen diameter.

Data Analysis Whenever discrepancies were noted between conventional invasive coronary angiography and CT coronary angiography results, a consensus conference was held among reviewers of both techniques to verify that identified coronary lesions were allocated to the same coronary segments by both techniques. We correlated the heart rate and the performance of CT coronary angiography in these patients. Calculations were performed using a heart rate of 80 beats per minute (bpm) or greater as a cutoff value. We also correlated the total calcium score and the performance of CT coronary angiography in these patients. Calculations were performed using a score of 400 or more (Agatston score equivalent) as a cutoff value.

Statistical Analysis Patient data were documented by the departments of diagnostic imaging and cardiology of our institution. Categoric data were presented with absolute frequencies and percentages and were compared using the chi-square test; p values equal to or less than 0.01 were considered to be relevant. The distributions of continuous variables were presented with their mean and SD. Computations were performed using SPSS for Windows (version 12, Statistical Package for the Social Sciences).

Results CT coronary angiography was technically successful in 59 patients (89%). The average heart rate was 71 ± 11 bpm (range, 51–96 bpm). In seven patients (11%) the image quality was severely degraded because of multiple motion artifacts. The motion artifacts resulted from respiratory motion (n = 3 patients, 4.5%), cardiac arrhythmia (n = 3, 4.5%), and absent cardiac gating (n = 1, 2%) (Fig. 1). These studies were subsequently classified as nondiagnostic on CT coronary angiography and were excluded from the statistical analysis. In the remaining 59 patients, only 20 (3.1%) of 649 coronary segments were classified as nonevaluable on CT coronary angiography. After analyzing all coronary segments of 2 mm or more in diameter that were shown adequately on both invasive coronary angiography and CT coronary angiography (a total of 629 segments), the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT coronary angiography for identifying significant CAD were

AJR:186, January 2006

MDCT Versus Invasive Coronary Angiography in Chest Pain

A

B

Fig. 1—Two patients with unstable angina pectoris. A, 45-year-old man. Stepladder artifacts caused by respiratory motion. Sagittal maximum-intensity-projection image shows stepladder artifacts affecting both anterior chest wall (solid arrows) and cardiac chambers and great vessels (dotted arrows). B, 70-year-old woman. Stepladder artifacts caused by cardiac motion resulting from faulty ECG gating. Sagittal maximum-intensity-projection image shows stepladder artifacts affecting cardiac chambers and main pulmonary artery (dotted arrows) but not chest wall structures (solid arrow).

80% (68/85), 89% (482/544), 52% (68/130), 97% (482/499), and 87% (550/629), respectively. The overall accuracy for the left main, left anterior descending, left circumflex, and right coronary arteries was 93%, 88%, 86%, and 86%, respectively (Figs. 2 and 3). Of 17 undetected significant stenoses (20%; 17/85) on CT coronary angiography, seven (8.2%; 7/85) were borderline on CT coronary angiography (40–49% diameter stenoses). Of 62 false-positive assessments (11.4%; 62/544) on CT coronary angiography, 21 (3.9%; 21/544) were borderline (40–49% diameter stenoses) on invasive coronary angiography. When analyzing the present data with respect to correct diagnosis per patient, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CT coronary angiography for identifying significant CAD were 83% (29/35), 54% (13/24), 73% (29/40), 68% (13/19), and 71% (42/59), respectively. When assessing the effects of heart rate with respect to correct diagnosis per patient, we found an accuracy of 66% for CT coronary angiography for identifying significant CAD in the 44 patients with heart rates slower than 80 bpm and an accuracy of 87% for CT coronary angiography for identifying significant CAD in the 15 patients with heart rates faster than 80 bpm (p = not significant). The average Agatston score was 270 ± 358 (range, 0–1,488). When assessing the effects of calcium burden and Agatston score with re-

AJR:186, January 2006

spect to the correct diagnosis per patient, in the 29 patients with available calcium scores we found an accuracy of 77% for CT coronary angiography for identifying significant CAD in the 22 patients with calcium scores of less than 400, and an accuracy of 100% for CT coronary angiography for identifying significant CAD in the seven patients with calcium score greater than 400 (p = not significant). A total of 27 coronary stents were studied. Twenty-five stents were correctly identified as patent, one stent was correctly identified as having significant in-stent luminal narrowing, and one patent stent was not identified as a coronary stent because of extensive circumferential coronary artery wall calcification. Nevertheless, the corresponding arterial segment was correctly diagnosed as patent (Figs. 4 and 5). Discussion More than 1 million inpatient diagnostic cardiac catheterizations are performed annually in the United States [2]. Approximately 40% are estimated to be solely for diagnostic purposes [6]. Hence, an imperative need exists for a reliable noninvasive coronary artery imaging technique to reduce the number of diagnostic cardiac catheterizations and to triage therapeutic coronary interventions. Until recently, electron beam CT and MRI were the only imaging techniques explored for this purpose [10, 11]. Since the advent of MDCT technology with its superior spatial, contrast, and especially temporal resolutions, several investigators have

evaluated this technology for coronary imaging. Achenbach et al. [1] and Knez et al. [2] compared coronary artery scanning using 4MDCT scanners with invasive coronary angiography and reported similar negative predictive values of 96% and 98%. However, the percentages of nonevaluable coronary segments and arteries were 6% and 32%, respectively [1, 2]. To date, to our knowledge, only three studies comparing 16-MDCT scanners with invasive coronary angiography have been published. Nieman et al. [3] reported sensitivity, specificity, positive predictive value, and negative predictive value of 95%, 86%, 80%, and 97%, respectively, for identifying stenosis of 50% or more while using only 12 of the 16 detectors because of technical limitations. When considering that no segments were excluded from the analysis, their results are excellent. More recently, Ropers et al. [4] reported sensitivity, specificity, positive predictive value, and negative predictive value of 92%, 93%, 79%, and 97%, respectively, for identifying stenosis of 50% or more using all 16 detectors. However, in their series, 12% of the arterial segments were excluded from the statistical analysis after they were considered nonevaluable. Mollet et al. [5], in their recently published large series of 128 patients using a 16MDCT scanner, reported comparable results. Nevertheless, all those studies share two important characteristics: First, the study populations were composed of elective patients with suspected CAD, whereas patients with acute

179

Ghersin et al.

A

C

B

D

Fig. 2—72-year-old man with unstable angina pectoris. A, Volume-rendered CT angiogram shows stenotic lesion (arrow) in mid left anterior descending coronary. B and C, Corresponding curved planar reformations in two projections (B) and global view of coronary vessel tree (C) show several nonsignificant calcified plaques in left main and proximal left anterior descending (LAD) coronary arteries (dotted arrows) and significant heterogeneous plaque in mid left anterior descending coronary artery (solid arrows) that is causing 60% diameter stenosis. LMCA = left main coronary artery, CRX = circumflex coronary artery, RCA = right coronary artery. D, Corresponding conventional coronary artery catheterization, performed on following day, confirms nonsignificant stenotic lesions in left main and proximal left anterior descending coronary arteries (dotted arrows) and 62% diameter stenosis in mid left anterior descending coronary artery (solid arrow).

coronary syndromes such as unstable angina pectoris were excluded; and second, premedication with β–blockers was routinely used for lowering heart rate, resulting in average heart rates of 56, 62, and 58 bpm, respectively [3–5]. After the encouraging results of Nieman et al., Ropers et al., and Mollet et al. [3–5], our

180

main objective was to further evaluate and potentially extend the usefulness of the 16MDCT scanner technology by applying it to a unique study group composed of patients hospitalized for acute chest pain syndrome, most (89%) of whom were ultimately diagnosed as having an acute coronary syndrome. In addi-

tion, unlike prior series, we limited our exclusion criteria to only allergy to iodine contrast media, renal insufficiency, and arrhythmia. Prior cardiac surgery, coronary arterial interventions, and Agatston calcium scores did not serve as exclusion criteria. Furthermore, we elected not to use routine premedication with

AJR:186, January 2006

MDCT Versus Invasive Coronary Angiography in Chest Pain

A

B

C

D

Fig. 3—45-year-old man with anterior wall myocardial infarction. A, Volume-rendered CT angiogram shows stenotic lesion in mid left anterior descending coronary artery (arrow). B and C, Corresponding curved planar reformations in two orthogonal projections (B) and global view of coronary vessel tree (C) show single soft plaque (solid black arrow) in mid left anterior descending (LAD) coronary artery that is causing 70% diameter stenosis. LMCA = left main coronary artery, CRX = circumflex coronary, RCA = right coronary artery. D, Corresponding conventional coronary catheterization, performed 5 days later, confirms 80% diameter stenosis (arrow) in mid left anterior descending coronary artery.

β–blockers for lowering the heart rate, and indeed our average heart rate was 71 ± 11 bpm (range, 51–96 bpm). Keeping in mind the unique characteristics of our study group and

AJR:186, January 2006

study CT protocol, our results are encouraging. We obtained sensitivity, specificity, positive predictive values, and negative predictive values of 80%, 89%, 52%, and 97%, respectively.

These values are comparable to the results of Nieman et al., Ropers et al., and Mollet et al., except for a mildly reduced sensitivity and a significantly lower positive predictive value.

181

Ghersin et al. Fig. 4—57-year-old man 2 days after percutaneous transluminal coronary angioplasty and stent placement in proximal left anterior descending artery. A, Volume-rendered CT angiogram shows stent in proximal left anterior descending artery (thick black arrow). Note protrusion of proximal end of stent (dotted white arrow) into proximal left circumflex artery (thin black arrow). Also note left main coronary artery (solid white arrow). B and C, Corresponding curved planar reformation (B) and conventional coronary catheterization (C) show patent stent lumen (solid arrows). Proximal left circumflex artery has significant stenosis (dotted arrows) caused by protrusion of proximal end of stent.

A

B These differences may be attributed to the distinctive qualities of our study group; for example, a large fraction of acutely ill patients had elevated heart rates and elevated calcium scores that caused motion artifacts and calcium plaque oversizing, respectively [6]. As a result,

182

we had a substantial number of false-positive interpretations, which led to a relatively low positive predictive value. We studied the correlation between heart rates and the performance of CT coronary angiography with respect to correct diagnosis

C per patient in detecting significant CAD. We found similar performance for CT coronary angiography in terms of accuracy (66% and 87%) for patients with heart rates of less than and greater than 80 bpm, respectively. Thus, our results both call into question the routine

AJR:186, January 2006

MDCT Versus Invasive Coronary Angiography in Chest Pain

A

B

Fig. 5—74-year-old man 2 years after percutaneous transluminal coronary angioplasty and stent placement in proximal left anterior descending artery. A, Curved planar reformation shows hypodense filling defects in both ends of stent (solid arrows) and faint visualization of patent stent lumen (dotted arrow). B, Corresponding conventional coronary catheterization shows string-like, significantly narrowed residual lumen (dotted arrow).

need for pharmacologic manipulation in most patients and raise the possibility that betablockade can potentially improve results in a small minority of cases. Possibly a synergistic deteriorating effect exists between motion artifacts (induced by cardiac motion caused by elevated heart rates and respiratory motion) and calcium plaque oversizing (induced by elevated calcium scores). Interactions between these two factors when they are simultaneously present in the same coronary lesion, and the possibility that beta-blockade might potentially improve results in such cases, require further study. We also studied the correlation between the arterial calcium burden, as reflected by the total calcium score (Agatston score equivalent), and the performance of CT coronary angiography with respect to the correct diagnosis per patient in detecting significant CAD. We found similar performance of CT coronary angiography in terms of accuracy for patients with calcium scores less than 400 (accuracy, 77%) and for patients with calcium scores greater than 400 (accuracy, 100%). Our results are significantly better than the results reported by Kuettner et al. [12]. In their study, using a 4-MDCT scanner, a calcium score threshold of 335 or less (Agatston score equivalent) resulted in an overall accuracy of 56%. Our study shows that current 16-MDCT technology may be substantially superior to 4-MDCT technology and can partially overcome the limitations induced by calcium

AJR:186, January 2006

oversizing or blooming, allowing reasonable performance even with markedly elevated calcium scores. Of course, newer improved MDCT technology using 40–64 detectors and offering improved spatial and temporal resolutions may provide a true noninvasive substitute to conventional invasive coronary angiography and intravascular sonography by enabling a true 3D “lumenogram” of the coronary arteries, combined with plaque detection and characterization [6]. We obtained 100% sensitivity in identifying coronary stent patency and significant instent luminal stenosis by direct visualization of the stent lumen. In comparison, Maintz et al. [7] found 20 of 47 evaluated stents were nonevaluable on a 4-MDCT scanner because of insufficient lumen visibility. Our improved results are attributed to the improved spatial, contrast, and temporal resolutions of 16MDCT scanners and to automated and semiautomated reconstruction algorithms offering curved multiplanar reformations through the stent lumen. In addition, window level adjustments were helpful in differentiating stent walls from stent lumen. The potential limitations of our study are our relatively small study group and the fact that we used conventional invasive coronary angiography, with measurement of lesion severity by electronic calipers, as the gold standard, instead of quantitative coronary angiography. The cardiology literature states that visual estimates of lesion severity are consis-

tently and significantly higher than quantitative measurements, and it advocates the use of quantitative coronary angiography as a better reflector of the actual disease burden [13]. Nevertheless, in most clinical catheterization laboratories, quantitative coronary angiography is still a purely investigational tool, whereas conventional invasive coronary angiography, with visual estimates of lesion severity, routinely guides therapeutic coronary interventions. As an adequate substitute to quantitative coronary angiography, we decided to measure lesion severity with electronic calipers as our gold standard. Our decision is supported by the findings of Uehata et al. [14], who concluded that electronic calipers are a rapid, convenient, and comparable alternative to quantitative coronary angiography. Taking into account these limitations, our study illustrates the improved capabilities of 16-MDCT scanners in showing clinically relevant coronary artery disease in patients with acute chest pain syndrome and relatively increased calcium scores. Furthermore, despite prior concerns, our results show relative high sensitivity and specificity, even in the setting of normal heart rates. Although larger, double-blinded, multicenter studies are required to validate our results, it is likely that in the near future CT coronary angiography may play an important role in the management and triage of patients with acute chest pain syndrome presenting to the emergency department. As a result, the percentage of diagnostic cardiac catheterizations

183

Ghersin et al. performed in these patients could be significantly reduced. Acknowledgments We appreciate the thoughtful insights and assistance in manuscript preparation and review provided by the BIDMC/Rambam Medical Center–Boston/Haifa Research Partnership Initiative, under the direction of S. Nahum Goldberg. We also acknowledge the valuable contribution of the CT team of radiographers in the Department of Diagnostic Imaging, Rambam Medical Center, Haifa, Israel, with special thanks to Shmuel Weizman, Pesah Ladovich, and Chava Wulf.

References 1. Achenbach S, Giesler T, Ropers D, et al. Detection of coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically-gated, multislice spiral computed tomography. Circulation 2001; 103:2535–2538 2. Knez A, Becker CR, Leber A, et al. Usefulness of multislice spiral computed tomography angiography for determination of coronary artery stenoses. Am J Cardiol 2001; 88:1191–1194

184

3. Nieman K, Cademartiri F, Lemos PA, Raaijmakers R, Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2002; 106:2051–2054 4. Ropers D, Baum U, Pohle K, et al. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003; 107:664–666 5. Mollet NR, Cademartiri F, Nieman K, et al. Multislice spiral computed tomography coronary angiography in patients with stable angina pectoris. J Am Coll Cardiol 2004; 43:2265–2270 6. Hoffmann MH, Shi H, Schmid FT, Gelman H, Brambs HJ, Aschoff AJ. Noninvasive coronary imaging with MDCT in comparison to invasive conventional coronary angiography: a fast-developing technology. AJR 2004; 182:601–608 7. Maintz D, Grude M, Fallenberg EM, Heindel W, Fischbach R. Assessment of coronary arterial stents by multislice-CT angiography. Acta Radiol 2003; 44:597–603 8. Austen WG, Edwards JE, Frye RL, et al. A reporting system on patients evaluated for coronary artery disease. Report of the Ad Hoc Committee for Grading of Coronary Artery Disease, Council on Cardiovascular Surgery, American Heart Association.

Circulation 1975; 51[4 suppl]:5–40 9. Baptista J, Arnese M, Roelandt JR, et al. Quantitative coronary angiography in the estimation of the functional significance of coronary stenosis: correlations with dobutamine-atropine stress test. J Am Coll Cardiol 1994; 23:1434–1439 10. Achenbach S, Moshage W, Ropers D, Bachmann K. Curved multiplanar reconstructions for the evaluation of contrast-enhanced electron beam CT of the coronary arteries. AJR 1998; 170:895–899 11. Kim WY, Danias PG, Stuber M, et al. Coronary magnetic resonance angiography for the detection of coronary stenoses. N Engl J Med 2001; 345:1863–1869 12. Kuettner A, Kopp AF, Schroeder S, et al. Diagnostic accuracy of multidetector computed tomography coronary angiography in patients with angiographically proven coronary artery disease. J Am Coll Cardiol 2004; 43:831–839 13. Goldberg RK, Kleiman NS, Minor ST, Abukhalil J, Raizner AE. Comparison of quantitative coronary angiography to visual estimates of lesion severity pre and post PTCA. Am Heart J 1990; 119:178–184 14. Uehata A, Matsuguchi T, Bittl JA, et al. Accuracy of electronic digital calipers compared with quantitative angiography in measuring coronary arterial diameter. Circulation 1993; 88(4 pt 1):1724–1729

AJR:186, January 2006