December 2008, VOLUME 191
NUMBER 6

Recommend & Share

December 2008, Volume 191, Number 6

Cardiopulmonary Imaging

Original Research

Meta-Analysis of 40- and 64-MDCT Angiography for Assessing Coronary Artery Stenosis

+ Affiliation:
1All authors: Department of Radiology, C2-S, Leiden University Medical Center, Albinusdreef 2, 2333 ZA Leiden, The Netherlands.

Citation: American Journal of Roentgenology. 2008;191: 1667-1675. 10.2214/AJR.07.4022

ABSTRACT
Next section

OBJECTIVE. The purpose of our study was to assess the diagnostic performance of thin-slice (≤ 0.625 mm) MDCT coronary angiography compared with invasive coronary angiography for the detection of significant (≥ 50%) stenosis.

MATERIALS AND METHODS. Twenty-two articles on 40- and 64-MDCT coronary angiography were included. Sensitivity and specificity were calculated on a per-patient and per-segment basis; in addition, proximal versus distal segments were evaluated. The effect of nonevaluable patients, nonevaluable segments, and disease prevalence on diagnostic performance was assessed.

RESULTS. Pooled sensitivity on a patient level was 97.7% ([95% CI] 96.2–98.7%) and specificity 91.0% (88.5–93.1%). Pooled sensitivity on a segmental level was 90.8% (89.0–92.4%) and specificity 95.7% (95.2–96.1%); for proximal segments, respectively, 94.2% (92.3–95.7%) and 94.1% (93.4–94.8%), and for distal segments 84.8% (81.1–88.0%) and 96.9% (96.4–97.4%). If nonevaluable MDCT investigations were included, the per-patient specificity was reduced from 91.0% to 89.1% (p > 0.05) when allocating excluded patients as having significant coronary artery stenosis, and the sensitivity was reduced from 97.7% to 96.2% (p > 0.05) when allocating excluded patients as not having significant stenosis. The per-patient prevalence of coronary artery stenosis had no significant influence on the sensitivity for detecting significant stenosis.

CONCLUSION. Forty- and 64-MDCT provide good-to-excellent performance in detecting or ruling out significant coronary artery stenosis, with better results for proximal than for distal coronary artery segments.

Keywords: cardiac, coronary angiography, CT, diagnostic imaging, stenoses

Introduction
Previous sectionNext section

Although catheter-based invasive coronary angiography is still considered the standard of reference for evaluating coronary artery stenoses, the development of MDCT angiography (CTA) for routine noninvasive clinical evaluation of the coronary arteries has high potential for well-defined patient populations. Several meta-analyses have assessed the value of 4-, 16-, and 64-MDCT coronary angiography for diagnosing coronary artery disease (CAD) [16]. These studies showed that diagnostic performance for evaluating coronary artery stenoses was limited for 4- and 16-CTA [1, 2, 4, 6]; however, considerable improvements were reported for thin-slice 64-CTA [1, 4, 6, 7]. Inclusion of 64-CTA articles within these meta-analyses is limited and analysis was restricted to per-patient or overall-segmental analysis [1, 4, 5, 7, 8].

Despite the increase in image quality with 64-MDCT compared with previous generation scanners, a substantial percentage (3–11%) of coronary artery segments have been reported nonevaluable due to image artifacts [913]. The sensitivities and specificities for detecting significant coronary artery stenoses with 64-CTA, using invasive coronary angiography as the standard of reference, have been reported as good-to-excellent. However, in a substantial number of 64-CTA investigations, patients or coronary artery segments that were nonevaluable had been excluded from analysis beforehand. Therefore, study results should be interpreted with care [14].

Coronary 40-CTA as well as coronary 64-CTA with 32 active detector rows and z-flying focal spot investigations have reported performance results similar to that of coronary 64-CTA with 64 active detector rows. Forty- and 64-MDCT scanners are capable of coronary angiography at submillimeter slice thickness (detector-width in the range 0.5–0.625 mm) [1519]. Also, these scanners allow short scanning times and consequently breath-hold times have been shortened to less than 15 seconds.

According to expert opinion, CTA is now considered an appropriate clinical imaging technique for detecting CAD in well-defined clinical contexts [20]. With the growing number of installed MDCT scanners, requests for noninvasive evaluation of the coronary arteries will likely increase. Recently, many new articles have been published that report on the diagnostic performance of 40- and 64-MDCT in patients with suspected CAD. This justifies reviewing systematically the reported diagnostic performance, while taking into account the effect of nonevaluable patients or coronary artery segments. Various articles have shown the performance of 40- and 64-MDCT in proximal versus distal segments.

Accordingly, the purpose of this study was to perform a meta-analysis of the reported diagnostic performance of 40- and 64-MDCT compared with invasive coronary angiography for detection of significant (≥ 50%) coronary artery stenoses on a per-patient as well as a per-segment basis, including an evaluation of proximal versus distal coronary artery segments and including in the analysis a correction for nonevaluable patients and segments.

Materials and Methods
Previous sectionNext section
Data Selection

Publications were searched in PubMed and Embase for the occurrence of the following search terms in the MeSH headings: ([diagnostic accuracy] OR [sensitivity AND specificity]) AND (CT) AND ([coronary stenosis] OR [coronary artery disease] OR [myocardial infarction]) OR ([unstable angina] OR [acute coronary syndrome]). The search was limited to peer-reviewed articles concerning human subjects and published in the English language. In addition, the articles were screened manually for references regarding the subject of the meta-analysis. The search of literature ranged from January 1, 2005, to December 15, 2007.

Studies were included in the meta-analysis if they met the following criteria: prospective or retrospective clinical trial study with imaging performed by a 40- or 64-MDCT scanner with detector-element width ≤ 0.625 mm; imaging performed in patients suspected for CAD; a significant coronary artery stenosis was graded as a ≥ 50% diameter reduction; invasive coronary angiography was used as standard of reference independent of CTA findings; reported data included the absolute numbers of true-positive, true-negative, false-positive, and false-negative findings or allowed recalculation of 2 × 2 contingency tables; and analysis was performed on a per-patient or per-segment level, in which significant CAD on the patient level was defined as having at least one significant stenosis in at least one segment.

Studies were excluded based on the following criteria: overlapping of study samples, obvious patient selection bias (e.g., studies with inclusion of 50% or more patients with either known CAD, balloon angioplasty, stents, or coronary artery bypass graft [CABG]), studies that did not specify nonevaluable scans and segments, and studies with frequent extrasystole heartbeats (> 5 per minute). For the subanalysis, studies were excluded when segment analysis was not specified per segment or by proximal versus distal segments.

Data Extraction

Study data were extracted from the original publications independently by two observers and then in consensus if disagreement existed between the observers. We used the Standards for Reporting of Diagnostic Accuracy (STARD) criteria [21] as a guideline, in which 16 items that were of specific relevance to CAD and diagnostic imaging have been implemented in our study design. Extracted from each study were the authors, journal citation, and publication date. Study population characteristics included the sample size and number of subjects evaluated and excluded, number of patients or coronary artery segments that were nonevaluable, type of analysis by patient or segment; the segment model used in segment analysis, whether a minimum size or all segment parts were used for vessel evaluation, prevalence of CAD and multivessel disease, CT scanner used, detector width, sex, mean age, and mean heart rate.

When patients or coronary artery segments had been excluded from the studies after the CTA investigation had been made, these nonevaluable patients or nonevaluable coronary artery segments were taken into account by recalculations. In the recalculations, patients or segments were regarded as either having or not having significant stenosis. The absolute numbers of true-positive, true-negative, false-positive, and false-negative findings as well as uncertain results were retrieved or calculated and were defined as stated in Table 1.

TABLE 1: Categorization of Results on MDCT Angiography (CTA)

Data were extracted on a per-patient and a per-segment basis in which, on the segmental level, the outcome was calculated for proximal versus distal segments as well as for individual proximal coronary artery segments. In addition, data about the left main segment was extracted from the included articles with separate analysis on a vessel level.

As a means to convert the different segment classification systems used in the various publications included in the meta-analysis into a single segment classification system, the coronary artery segments of each included publication were first mapped to the extensive Bypass Angioplasty Revascularization Investigation (BARI) model [22]. This allowed converting the various segment classification systems to a uniform model. Then the coronary artery segments were indexed as shown in Table 2. Categorization into proximal versus distal segments followed the American College of Cardiology/American Heart Association (ACC/AHA) 15-segment model [23]. Six segments were assigned as proximal segments (Table 2) and in addition, ≥ 2-mm-sized coronary artery segments were also included in the proximal segment group. The remainder of indexed segments, as well as < 2-mm-sized segments, were assigned to the distal segment group.

TABLE 2: Uniform Indexation of Coronary Artery Segments for Each Included Publication

Statistical Analysis

The meta-analysis was performed using the random effects model (DerSimonian-Laird) at a confidence level of 0.95 (p value of 0.05). Sensitivities and specificities were pooled by weighted averages of the sample sizes. The pooled sensitivities, specificities, and CIs were calculated on per-patient and on per-segment levels. The analysis, in addition, included correction for nonevaluable segments (Table 1). The heterogeneity was assessed using the Cochrane I2 (chisquare) test.

The summary receiver operating characteristic (SROC) curve contains points (1 — specificity, sensitivity) of different studies. Symmetric SROC curves were determined with the corresponding area under the curve (AUC) and Q point (the point on the curve where sensitivity equals specificity). A value of 0.5 was added to studies containing zero values. The higher the curve above the diagonal 45° line, the higher the diagnostic accuracy of the test. All the above statistics were calculated using Meta-Disk, version 1.4 (open source software) [24], a well-documented and validated software program in which a variety of statistical techniques commonly used in meta-analyses are implemented. Dependence of the sensitivity on clinical variables was evaluated by linear regression with the sensitivity as the dependent variable in SPSS, version 15.0, statistical software. The sensitivity is considered to be independent of a clinical variable if the value of the regression coefficient is insignificant.

On the per-patient level, the following analyses were performed: sensitivity and specificity with calculations for nonevaluable scans as either excluded, considered positive, or considered negative. Also, regression analysis of sensitivity versus disease prevalence and sensitivity against heart rate (excluding nonevaluable patients) was performed.

Additionally, on the per-segment level the sensitivity and specificity were calculated with nonevaluable segments as either excluded, considered positive, or considered negative. These calculations were performed for proximal and distal segments separately. For comparison with accuracy values in existing literature, all segments were combined for analysis (excluding nonevaluable segments).

figure
View larger version (24K)
Fig. 1 Flow diagram shows reviewing selection process.

Results
Previous sectionNext section
Study Description

The search resulted in 823 articles. After reading the title or abstract, 109 articles remained for further evaluation. From these articles, 22 studies met the study criteria and were included for analysis [9, 1619, 2543]. In addition, one article with a comparison of ≥ 2 versus < 2 mm segments was included in the proximal versus distal segment analysis [42]. Cross-referencing was performed on included articles and previous meta-analyses on 40- and 64-MDCT and yielded four more articles. The systematic review selection process and reasons for excluding studies are summarized in Figure 1. For the two studies that included patients with CABG, the native vessels were included in analysis. Ten studies were included that also contained patients with stents. The stented segments were not included in the analysis.

Table 3 summarizes general information and Table 4 contains extracted accuracy data for the included studies. A total of 1,331 patients and 10,561 segments were analyzed (Table 3). The percentage of patients excluded from analysis after CT had been performed ranged from 0% to 13.5% (median, 0%). In total, 1,307 (98.2%) patients were scanned with diagnostic quality. The percentage of nonevaluable proximal and distal coronary artery segments ranged from 0% to 15.4% (median, 2.0%) and from 0% to 19.3% (median, 2.7%), respectively. The mean age was 60 years (age range, 54–69 years), and the mean heart rate was 64 beats per minute (bpm) (range, 58–73 bpm). The prevalence of CAD on a patient level ranged from 18% to 85% (mean, 50%) (Table 3). The detailed accuracy data listed in Table 4 refers to patients and, in addition, to the proximal and distal coronary artery segments.

TABLE 3: General Information for the Studies Included in the Meta-Analysis

TABLE 4: Overview of the Data on Accuracy of Coronary MDCT Angiography (in Absolute Numbers) from the Articles in the Meta-Analysis

Diagnostic Performance of 40- and 64-MDCT

The diagnostic performance of 40- and 64-MDCT for the detection of significant (≥ 50%) coronary artery stenoses on a patient level and on a segmental level by evaluation of proximal versus distal segments, as well as for individual proximal coronary artery segments, is summarized in Table 5; nonevaluable patients or segments were excluded from the analysis. The SROC curves for the analysis on a patient level and for proximal and distal segments are plotted in Figures 2A, 2B, and 2C. Sensitivity on a per-patient basis was significantly higher than on a segmental level, whereas specificity was higher on a segmental level (p < 0.05). On the patient level, sensitivity and specificity for detecting significant coronary artery stenosis were 97.7% (96.2–98.7%) and 91.0% (88.5–93.1%), respectively. Heterogeneity was low (I2 = 0.0% for sensitivity and 23% for specificity).

TABLE 5: Calculated Sensitivity and Specificity for MDCT Compared with Invasive Coronary Angiography with No Correction Applied

On the segmental level, sensitivity was significantly higher in proximal compared with distal segments (94.2% vs 84.8%, p < 0.05), whereas specificity was significantly lower in proximal compared with distal segments (94.1% vs 96.9%, p < 0.05). Heterogeneity was rather high for these analyses (I2 = 56.0% vs 85.0% for sensitivity and 89.7% vs 86.0% for specificity).

Diagnostic Performance with Correction for Nonevaluable Patients and Nonevaluable Coronary Artery Segments

Calculations were repeated taking into account corrections for MDCT examinations that yielded nonevaluable patients or nonevaluable segments. Table 6 shows the sensitivities and specificities of the CTA investigations calculated when these nonevaluable patients or coronary artery segments would have been included and considered as having significant stenosis. With this correction, the sensitivity and specificity on a patient level were 97.7% and 89.1%, respectively; differences were not statistically significant. On a segmental level, the sensitivity and specificity for detecting significant coronary artery stenosis in proximal segments were, respectively, 94.4% and 91.6%; the decrease in specificity was statistically significant (p < 0.05). For distal segments, sensitivity and specificity were, respectively, 85.8% and 91.7%; the decrease in specificity was statistically significant (p < 0.05).

TABLE 6: Calculated Sensitivity and Specificity for MDCT Compared with Invasive Coronary Angiography Where Nonevaluable Patients or Segments Are Included as Having Significant Artery Stenosis

Table 7 shows the results when nonevaluable scans are corrected for by including them as negative results. The sensitivity and specificity on the patient level are then 96.2% and 91.2%, respectively; the changes are not statistically significant. On the segmental level, the sensitivity and specificity for detecting significant coronary artery stenosis in proximal segments were 90.1% and 94.3%; the decrease in sensitivity was statistically significant (p < 0.05). For distal segments, sensitivity and specificity were 78.9% and 97.1%, with a statistically significant decrease in sensitivity (p < 0.05).

TABLE 7: Calculated Sensitivity and Specificity for MDCT Compared with Invasive Coronary Angiography Where Nonevaluable Patients or Segments Are Included as Not Having Significant Artery Stenosis

figure
View larger version (17K)
Fig. 2A Summary receiver operating characteristic (SROC) curves for analysis on patient level and for proximal and distal segments. SROC curves for analysis of diagnostic accuracy of 40- and 64-MDCT coronary angiography on patient level (A), proximal segmental level (B), and distal segmental level (C). SROC curve provides graphic display of diagnostic accuracy by plotting sensitivity versus one minus specificity and accounts for differences in diagnostic threshold among studies. Area under the curve (AUC) and Q point (point on curve where sensitivity equals specificity) are also given, including standard error (SE).

figure
View larger version (17K)
Fig. 2B Summary receiver operating characteristic (SROC) curves for analysis on patient level and for proximal and distal segments. SROC curves for analysis of diagnostic accuracy of 40- and 64-MDCT coronary angiography on patient level (A), proximal segmental level (B), and distal segmental level (C). SROC curve provides graphic display of diagnostic accuracy by plotting sensitivity versus one minus specificity and accounts for differences in diagnostic threshold among studies. Area under the curve (AUC) and Q point (point on curve where sensitivity equals specificity) are also given, including standard error (SE).

figure
View larger version (17K)
Fig. 2C Summary receiver operating characteristic (SROC) curves for analysis on patient level and for proximal and distal segments. SROC curves for analysis of diagnostic accuracy of 40- and 64-MDCT coronary angiography on patient level (A), proximal segmental level (B), and distal segmental level (C). SROC curve provides graphic display of diagnostic accuracy by plotting sensitivity versus one minus specificity and accounts for differences in diagnostic threshold among studies. Area under the curve (AUC) and Q point (point on curve where sensitivity equals specificity) are also given, including standard error (SE).

Individual Proximal Coronary Artery Segment Evaluation

Table 5 also shows results for analyses of individual proximal segments, without correction for nonevaluable segments. The best sensitivity and specificity for assessing significant coronary artery stenosis were obtained in segment 1 (left main), whereas somewhat worse results were obtained in segment 11 (middle right coronary artery) and 7 (proximal circumflex artery).

Effect of Disease Prevalence and Heart Rate on Sensitivity

The regression coefficient for disease prevalence was 0.003 (—0.049 to 0.054), indicating that there is no statistically significant dependence of sensitivity for detecting ≥ 50% stenosis on disease prevalence. However, sensitivity in detecting ≥ 50% stenosis showed a statistically significant negative relation with heart rate, with a coefficient of —0.003 (—0.005 to —0.002), indicating a reduction in sensitivity of 0.3% per heartbeat per minute increase.

Discussion
Previous sectionNext section

This systematic review including 22 articles shows high diagnostic accuracy for the current generation of 40- and 64-MDCT scanners in detecting significant coronary artery stenosis in patients suspected of having CAD. Sensitivities and specificities were found to be good-to-excellent for evaluation on a per-patient as well as on a per-segment basis. Recalculation with correction for nonevaluable patient examinations or nonevaluable segments had a limited but statistically significant effect on diagnostic accuracy. Sensitivity was shown to be significantly better for proximal than for distal coronary artery segments. Disease prevalence had no influence on CTA sensitivity.

Recent meta-analyses on 16- and 64-MDCT have already reported improved performance of 64-MDCT compared with 16-MDCT [1, 5, 7]. We now show results of a larger amount of publications, in which we additionally investigated the effect of proximal versus distal coronary artery segments on accuracy as well as the effect of nonevaluable patients and coronary artery segments. The excellent sensitivity (98%) and good specificity (91%) on the patient level make 40- and 64-MDCT scanners with small (≤ 0.625 mm) detectors and fast-rotating x-ray tubes (rotation time < 0.5 second) excellent tools for detecting or ruling out significant coronary artery stenosis in a carefully selected subgroup of patients suspected of significant CAD. A high sensitivity on a patient level is of vital importance in ruling out disease; the lower the sensitivity, the more often disease will be missed. Specificity plays an important role in the reduction of the overall cost of the evaluation of suspected CAD. The higher the specificity, the fewer patients will undergo unnecessary invasive coronary angiography due to a false-positive finding.

In addition, separate meta-analysis for proximal (or ≥ 2.0 mm) versus distal (or < 2.0 mm) coronary artery segments was performed. Analyzing proximal versus distal segments is relevant because the location of stenosis may have impact on left ventricular (LV) function: On echocardiography stress testing, significantly greater wall motion abnormalities have been found among patients with proximal than among those with middle and distal stenosis [44]. Also, in patients after left anterior descending coronary artery myocardial infarction, LV functional recovery has been found delayed and the LV chamber enlarged in patients with proximal lesions, whereas LV function was found preserved among patients with distal lesions [45]. Coronary artery lumen diameters become significantly smaller from their proximal to their distal parts [46]. For 64-MDCT, it has been found that the image quality of segments larger than 2.0 mm is better than for segments smaller than 2.0 mm, although more than half of the segments are smaller than 2.0 mm [12]. We have now shown that the sensitivity for finding coronary artery stenosis is statistically significantly better for proximal than for distal coronary artery segments. Thus, stenoses in the larger proximal parts of the coronary arteries that may have greater clinical impact are better detected than those in the distal smaller segments. Furthermore, we performed additional meta-analysis including nonevaluable patients and coronary artery segments as having or not having coronary artery stenosis.

Several studies excluded nonevaluable patients or segments, which may potentially cause apparently better results than had these patients or segments been included. In the articles included, only a few examinations on a patient level were excluded, and including these examinations in analysis as having or not having coronary artery stenosis had no significant effect on the outcome of this meta-analysis. However, in articles that report on evaluability of segments, the number of excluded segments (median, 2.0% for proximal and 2.7% for distal segments) had a statistically significant effect on outcome if these segments were included in the analysis. Including nonevaluable segments as positive resulted in a significant reduction of specificity, whereas including them as negative (for distal segments) resulted in a significant reduction of sensitivity. This can be explained by the number of nonevaluable patients and segments with and without known disease added to the number of true-positive and false-positive findings, respectively. Because the nonevaluable patients or segments are, on average, equally distributed over patients or segments with or without disease (because CAD prevalence in the assessable part of the patient group is, on average, 50%), adding these to the false-positive findings has more impact than on the true-positive findings, hence a more stable sensitivity. For the strategy of including nonevaluable scan results as negative, an analogue reasoning holds true.

On a segmental level, sensitivity was found lower and specificity higher compared with the patient analysis. This is expected because patients are defined as having CAD if any of the segments has a significant stenosis. In patients with multivessel CAD, the likelihood of missing all stenoses is lower than of missing just one stenosis, hence a higher sensitivity. On the other hand, specificity can be expected to be lower on a patient level than on a segmental level because on a per-patient basis, many segments have the potential to provide a false-positive finding. At least one false-positive finding in all of a patient's segments is more likely than one false-positive finding in a single segment, hence a lower patient specificity.

We found that our fairly homogeneous patient population also exhibited significant heterogeneity on a segmental level, when analysis was repeated with only the eight studies that reported both patient and segmental analyses (p = 0.54 vs p = 0.00 for heterogeneity in sensitivity on a patient vs all-segments analysis, respectively, and p = 0.24 vs p = 0.00 for heterogeneity in specificity). One possible reason for this is the variation in the diseased vessels in combination with considerable differences in prevalence of multivessel disease across the included trials (Table 3). Deriving the per-patient level from the per-segment level reduces this heterogeneity. Also, there were fewer studies on a segmental than a patient level (11 vs 20). The SROC curves in Figures 2A, 2B, and 2C of the per-patient and per-segment analyses, however, bear high similarity and can be viewed as showing consistency.

Evaluation of proximal segments showed the best outcome for the left main segment (sensitivity of 100% and specificity of 99.1%) and the worst outcome for the middle segment of the right coronary artery (sensitivity of 81.3% and specificity of 95.1%). These results can be explained by vessel size and motion. The left main coronary artery has the largest diameter of the coronary arteries. A large vessel means that a corresponding ≥ 50% stenosis is also large and therefore accuracy is less affected by a limited spatial resolution. The relatively poor results for the right coronary artery middle segment can be explained by the greater motion causing artifacts that reduce accuracy compared with the other coronary arteries [10, 47, 48]. In addition, our meta-analysis showed dependence of sensitivity for detecting significant coronary artery stenosis on heart rate. This is in line with findings of other authors who have shown an inverse relationship between image quality and heart rate [10, 38]. Image quality has been shown to be the major determinant regarding CT interpretation of coronary artery stenosis [14].

Prevalence is an important factor in the cost-effectiveness of CT in suspected CAD. Because a positive CT will often be followed by invasive coronary angiography, the added value of a strategy starting with CT is that a negative scan precludes an invasive coronary angiography. However, at high prevalence most CT scans will be positive, resulting in a higher overall diagnostic burden compared with a strategy of direct invasive coronary angiography. Therefore, current research focuses on the potential role of CT at intermediate and low pretest probabilities, mainly to rule out disease. It was found that sensitivity does not depend on CAD prevalence, which provides a part of the justification of the use of CT in patients with an intermediate likelihood of CAD. However, these results are based on studies with prevalence ranging between 15% and 85% and may not simply be extrapolated to low-prevalence patient groups.

Our study has some limitations. The majority of the studies included in the meta-analysis (17 of 22) were performed using equipment from a single manufacturer. This may have influenced the outcome and may limit general applicability of the results. Also, by design of the included studies that all used invasive coronary angiography as the standard of reference, the pretest probability (and with this the prevalence of CAD) was relatively high. Therefore, the study outcome may not be extrapolated to a low-to-intermediate pretest probability risk population, whereas low-to-intermediate pretest probability patients are more expected to be referred for CTA than high-risk patients. More studies are needed before conclusions about the performance of CTA in low-prevalence groups can be made.

To correct for nonevaluable patients or segments, for those with an assessable but unspecified invasive coronary angiography result, we made the assumption that the prevalence of CAD is the same as in the group with specified CT and an invasive coronary angiography result, that is, the occurrence of CT artifacts is assumed not to be correlated with prevalence of (an acute episode of) CAD. This assumption is supported by the fact that most artifacts are caused by motion [14] that has no obvious relation with CAD prevalence. However, in patients with nonevaluable segments caused by calcifications, disease prevalence might have been higher. Therefore, the analysis with correction for nonevaluable patients or segments should be regarded only as an estimation of trend.

In conclusion, this meta-analysis shows that current 40- and 64-MDCT scanners with their small detector rows (≤ 0.625 mm) have a good to excellent sensitivity and specificity for detecting or ruling out coronary artery stenoses in patients suspected of having CAD; the results are better for proximal coronary arteries. Disease prevalence had no influence on sensitivity.

This study was supported by the EC-EURATOM 6 Framework Programme (2002–2006) and forms part of the CT Safety and Efficacy (Safety and Efficacy of Computed Tomography [CT]: A Broad Perspective) project, contract FP/002388.

Address correspondence to L. J. M. Kroft ().

References
Previous sectionNext section
1. Sun Z, Jiang W. Diagnostic value of multislice computed tomography angiography in coronary artery disease: a meta-analysis. Eur J Radiol 2006; 60:279-286 [Google Scholar]
2. van der Zaag-Loonen HJ, Dikkers R, de Bock GH, Oudkerk M. The clinical value of a negative multidetector computed tomographic angiography in patients suspected of coronary artery disease: a meta-analysis. Eur Radiol 2006; 16:2748-2756 [Google Scholar]
3. Hamon M, Biondi-Zoccai GGL, Malagutti P, et al. Diagnostic performance of multislice spiral computed tomography of coronary arteries as compared with conventional invasive coronary angiography. J Am Coll Cardiol 2006; 48:1896-1910 [Google Scholar]
4. Stein PD, Beemath A, Kayali F, Skaf E, Sanchez J, Olson RE. Multidetector computed tomography for the diagnosis of coronary artery disease: a systematic review. Am J Med 2006; 119:203-216 [Google Scholar]
5. Vanhoenacker PK, Heijenbrok-Kal MH, Van Heste R, et al. Diagnostic performance of multidetector CT angiography for assessment of coronary artery disease: meta-analysis. Radiology 2007; 244:419-428 [Google Scholar]
6. Janne d'Othée B, Siebert U, Cury R, Jadvar H, Dunn EJ, Hoffmann U. A systematic review on diagnostic accuracy of CT-based detection of significant coronary artery disease. Eur J Radiol 2008; 65:449-461 [Google Scholar]
7. Hamon M, Morello R, Riddell JW, Hamon M. Coronary arteries: diagnostic performance of 16- versus 64-section spiral CT compared with invasive coronary angiography—meta-analysis. Radiology 2007; 245:720-731 [Google Scholar]
8. Sun Z, Lin CH, Davidson R, Dong C, Liao Y. Diagnostic value of 64-slice CT angiography in coronary artery disease: a systematic review. Eur J Radiol 2008; 67:78-84 [Google Scholar]
9. Nikolaou K, Knez A, Rist C, et al. Accuracy of 64-MDCT in the diagnosis of ischemic heart disease. AJR 2006; 187:111-117 [Abstract] [Google Scholar]
10. Ferencik M, Nomura CH, Maurovich-Horvat P, et al. Quantitative parameters of image quality in 64-slice computed tomography angiography of the coronary arteries. Eur J Radiol 2006; 57:373-379 [Google Scholar]
11. Wintersperger BJ, Nikolaou K, von Ziegler F, et al. Image quality, motion artifacts, and reconstruction timing of 64-slice coronary computed tomography angiography with 0.33-second rotation speed. Invest Radiol 2006; 41:436-442 [Google Scholar]
12. Pannu HK, Jacobs JE, Lai S, Fishman EK. Coronary CT angiography with 64-MDCT: assessment of vessel visibility. AJR 2006; 187:119-126 [Abstract] [Google Scholar]
13. Leshka S, Husmann L, Desbiolles LM, et al. Optimal image reconstruction intervals for noninvasive coronary angiography with 64-slice CT. Eur Radiol 2006; 16:1964-1972 [Google Scholar]
14. Kroft LJ, de Roos A, Geleijns J. Artifacts in ECG-synchronized MDCT coronary angiography. AJR 2007; 189:581-591 [Abstract] [Google Scholar]
15. Cordeiro MA, Miller JM, Schmidt A, et al. Noninvasive half millimetre 32 detector row computed tomography angiography accurately excludes significant stenoses in patients with advanced coronary artery disease and high calcium scores. Heart 2006; 92:589-597 [Google Scholar]
16. Grosse C, Globits S, Hergan K. Forty-slice spiral computed tomography of the coronary arteries: assessment of image quality and diagnostic accuracy in a non-selected patient population. Acta Radiol 2007; 48:36-44 [Google Scholar]
17. Lim MC, Wong TW, Yaneza LO, De Larrazabal C, Lau JK, Boey HK. Non-invasive detection of significant coronary artery disease with multisection computed tomography angiography in patients with suspected coronary artery disease. Clin Radiol 2006; 61:174-180 [Google Scholar]
18. Pouleur AC, le Polain de Waroux JB, Kefer J, et al. Usefulness of 40-slice multidetector row computed tomography to detect coronary disease in patients prior to cardiac valve surgery. Eur Radiol 2007; 17:3199-3207 [Google Scholar]
19. Watkins MW, Hesse B, Green CE, et al. Detection of coronary artery stenosis using 40-channel computed tomography with multisegment reconstruction. Am J Cardiol 2007; 99:175-181 [Google Scholar]
20. Hendel RC, Patel MR, Kramer CM, et al. ACCF/ACR/SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006 appropriateness criteria for cardiac computed tomography and cardiac magnetic resonance imaging: a report of the American College of Cardiology Foundation/American College of Radiology, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, American Society of Nuclear Cardiology, North American Society for Cardiac Imaging, Society for Cardiovascular Angiography and Interventions, and Society of Interventional Radiology. J Am Coll Cardiol 2006; 48:1475-1497 [Google Scholar]
21. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. AJR 2003; 181:51-55 [Abstract] [Google Scholar]
22. Alderman EL, Stadius M. The angiographic definitions of the Bypass Angioplasty Revascularization Investigation. Coron Artery Dis 1992; 3:1189-1207 [Google Scholar]
23. 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; 51:5-40 [Google Scholar]
24. Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6:31 [Google Scholar]
25. Fine JJ, Hopkins CB, Ruff N, Newton FC. Comparison of accuracy of 64-slice cardiovascular computed tomography with coronary angiography in patients with suspected coronary artery disease. Am J Cardiol 2006; 97:173-174 [Google Scholar]
26. Ghostine S, Caussin C, Daoud B, et al. Non-invasive detection of coronary artery disease in patients with left bundle branch block using 64-slice computed tomography. J Am Coll Cardiol 2006; 48:1929-1934 [Google Scholar]
27. Hacker M, Jakobs T, Hack N, et al. Combined use of 64-slice computed tomography angiography and gated myocardial perfusion SPECT for the detection of functionally relevant coronary artery stenoses. Nuklearmedizin 2007; 46:29-35 [Google Scholar]
28. Herzog C, Nguyen SA, Savino G, et al. Does two-segment image reconstruction at 64-section CT coronary angiography improve image quality and diagnostic accuracy? Radiology 2007; 244:121-129 [Google Scholar]
29. Herzog C, Zwerner PL, Doll JR, et al. Significant coronary artery stenosis: comparison on per-patient and per-vessel or per-segment basis at 64-section CT angiography. Radiology 2007; 244:112-120 [Google Scholar]
30. Leber AW, Johnson T, Becker A, et al. Diagnostic accuracy of dual-source multi-slice CT coronary angiography in patients with an intermediate pretest likelihood for coronary artery disease. Eur Heart J 2007; 28:2354-2360 [Google Scholar]
31. Meijboom WB, Mollet NR, Van Mieghem CA, et al. Pre-operative computed tomography coronary angiography to detect significant coronary artery disease in patients referred for cardiac valve surgery. J Am Coll Cardiol 2006; 48:1658-1665 [Google Scholar]
32. Meijboom WB, Mollet NR, Van Mieghem CA, et al. 64-slice CT coronary angiography in patients with non-ST elevation acute coronary syndrome. Heart 2007; 93:1325-1326 [Google Scholar]
33. Mollet NR, Cademartiri F, van Mieghem CA, et al. High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation 2005; 112:2318-2323 [Google Scholar]
34. Oncel D, Oncel G, Tastan A, Tamci B. Detection of significant coronary artery stenosis with 64-section MDCT angiography. Eur J Radiol 2007; 62:394-405 [Google Scholar]
35. Ong TK, Chin SP, Liew CK, et al. Accuracy of 64-row multidetector computed tomography in detecting coronary artery disease in 134 symptomatic patients: influence of calcification. Am Heart J 2006; 151:1323.e1-1323.e6 [Google Scholar]
36. Pugliese F, Mollet NR, Runza G, et al. Diagnostic accuracy of non-invasive 64-slice CT coronary angiography in patients with stable angina pectoris. Eur Radiol 2006; 16:575-582 [Google Scholar]
37. Pundziute G, Schuijf JD, Jukema JW, et al. Gender influence on the diagnostic accuracy of 64-slice multislice computed tomography coronary angiography for detection of obstructive coronary artery disease. Heart 2008; 94:48-52 [Google Scholar]
38. Raff GL, Gallagher MJ, O'Neill WW, Goldstein JA. Diagnostic accuracy of noninvasive coronary angiography using 64-slice spiral computed tomography. J Am Coll Cardiol 2005; 46:552-557 [Google Scholar]
39. Ropers D, Rixe J, Anders K, et al. Usefulness of multidetector row spiral computed tomography with 64- × 0.6-mm collimation and 330-ms rotation for the noninvasive detection of significant coronary artery stenoses. Am J Cardiol 2006; 97:343-348 [Google Scholar]
40. Scheffel H, Leschka S, Plass A, et al. Accuracy of 64-slice computed tomography for the preoperative detection of coronary artery disease in patients with chronic aortic regurgitation. Am J Cardiol 2007; 100:701-706 [Google Scholar]
41. Shabestari AA, Abdi S, Akhlaghpoor S, et al. Diagnostic performance of 64-channel multislice computed tomography in assessment of significant coronary artery disease in symptomatic subjects. Am J Cardiol 2007; 99:1656-1661 [Google Scholar]
42. Shapiro MD, Butler J, Rieber J, et al. Analytic approaches to establish the diagnostic accuracy of coronary computed tomography angiography as a tool for clinical decision making. Am J Cardiol 2007; 99:1122-1127 [Google Scholar]
43. Weustink AC, Meijboom WB, Mollet NR, et al. Reliable high-speed coronary computed tomography in symptomatic patients. J Am Coll Cardiol 2007; 50:786-794 [Google Scholar]
44. Tousoulis D, Rallidis L, Cokkinos P, Davies G, Nihoyannopoulos P. Relation between exercise and dobutamine stress-induced wall motion abnormalities and severity and location of stenosis in single-vessel coronary artery disease. Am Heart J 1999; 138:873-879 [Google Scholar]
45. Kachi E, Yoshino H, Watanuki A, Sasaki K, Sakata K, Ishikawa K. Effect of the stenosis location and severity on left ventricular function after single-vessel anterior wall myocardial infarction. Am Heart J 2001; 141:55-64 [Google Scholar]
46. Dodge JT, Brown BG, Bolson EL, Dodge HT. Lumen diameter of normal human coronary arteries: influence of age, sex, anatomic variation, and left ventricular hypertrophy or dilation. Circulation 1992; 86:232-246 [Google Scholar]
47. Johnson KR, Patel SJ, Whigham A, Hakim A, Pettigrew RI, Oshinski JN. Three-dimensional, time-resolved motion of the coronary arteries. J Cardiovasc Magn Reson 2004; 6:663-673 [Google Scholar]
48. Achenbach S, Ropers D, Holle J, Muschiol G, Daniel WG, Moshage W. In-plane coronary arterial motion velocity: measurement with electron-beam CT. Radiology 2000; 216:457-463 [Google Scholar]

Recommended Articles

Meta-Analysis of 40- and 64-MDCT Angiography for Assessing Coronary Artery Stenosis

Full Access, , , , ,
American Journal of Roentgenology. 2008;191:1676-1683. 10.2214/AJR.07.4026
Abstract | Full Text | PDF (735 KB) | PDF Plus (447 KB) 
Full Access, , , , , ,
American Journal of Roentgenology. 2008;191:1659-1666. 10.2214/AJR.07.3944
Abstract | Full Text | PDF (672 KB) | PDF Plus (369 KB) 
Full Access,
American Journal of Roentgenology. 2008;191:1698-1701. 10.2214/AJR.08.1193
Abstract | Full Text | PDF (818 KB) | PDF Plus (740 KB) 
Full Access, , , , , ,
American Journal of Roentgenology. 2008;191:1684-1689. 10.2214/AJR.07.4040
Abstract | Full Text | PDF (571 KB) | PDF Plus (270 KB) 
Full Access, , ,
American Journal of Roentgenology. 2009;192:574-583. 10.2214/AJR.08.1177
Abstract | Full Text | PDF (917 KB) | PDF Plus (867 KB) 
Full Access, , ,
American Journal of Roentgenology. 2009;192:584-598. 10.2214/AJR.08.1178
Abstract | Full Text | PDF (1107 KB) | PDF Plus (1133 KB) | Supplemental Material