Computed tomography for detecting coronary artery plaques: A meta-analysis

Atherosclerosis 219 (2011) 603–609

Contents lists available at SciVerse ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Computed tomography for detecting coronary artery plaques: A meta-analysis Dengfeng Gao a,∗,1 , Ning Ning b,c,1 , Youmin Guo c , Wende Ning d , Xiaolin Niu a , Jian Yang c,∗∗ a

Department of Cardiology, The Second Affiliated Hospital, Xi’an Jiaotong University School of Medicine, Xi’an, Shaanxi 710004, PR China Department of Nuclear medicine, The Second Affiliated Hospital, Xi’an Jiaotong University School of Medicine, Xi’an, Shaanxi 710004, PR China Department of Radiology, The First Affiliated Hospital, Xi’an Jiaotong University School of Medicine,Xi’an, Shaanxi 710061, PR China d Imaging center of Xi’an central hospital, Xi’an, Shaanxi 710004, PR China b c

a r t i c l e

i n f o

Article history: Received 28 June 2011 Received in revised form 3 August 2011 Accepted 10 August 2011 Available online 22 August 2011 Keywords: Computed tomography Intravascular ultrasound Plaque Meta-analysis

a b s t r a c t Objectives: CT is a novel noninvasive test for detection and analysis of coronary artery plaques. However, the ability of CT to detect and quantify coronary atherosclerotic plaque in vivo has never been systematically validated. We sought to conduct a meta-analysis to evaluate the accuracy of Computed tomography (CT) in detecting coronary artery plaques. Methods: We systematically searched the literature to identify reports published from 1966 up to January 2011 in Pubmed and EMBASE that described studies comparing CT and intravascular ultrasound (IVUS), the reference standard, in assessing coronary artery plaques. We sought reports that clearly indicated the number of true positive, false positive, false negative and true negative results. Results: We identified 17 reports of studies. On bivariate analysis, for CT diagnosis of any plaques, the mean sensitivity and specificity was 92% (95% confidence interval, 88–95%) and 93% (90–96%), respectively; for diagnosis of calcified plaque, 93% (84–97%) and 98% (96–99%); and for diagnosis of non-calcified plaque (soft or fibrotic), 88% (81–93%) and 92% (89–95%). Covariate analysis yielded a significantly higher sensitivity (95%) and specificity (94%) for CT scanners with more than 16 rows (P < 0.001) than for oldergeneration scanners (83% and 92%, respectively). Conclusion: CT should be considered the foremost noninvasive alternative to IVUS for detecting coronary artery plaques. Randomized studies at the patient level are needed to address the potential of CT for use in triage for altering management and outcomes in patients with suspected or onset coronary artery disease. © 2011 Elsevier Ireland Ltd. All rights reserved.

Coronary artery disease (CAD) continues to be one of the leading causes of morbidity and mortality around the globe. In the developed country, such as United States, it is estimated that approximately 16 million Americans would suffer from CAD [1] and a significant proportion of these are asymptomatic patients. CAD is a chronic progressive disease, and the detection of even minimal plaque burden across coronary vessels remains an important challenge for monitoring progression or regression of disease [2]. Intravascular ultrasound (IVUS) allows for cross-sectional imaging of coronary arteries and a comprehensive assessment of coronary atherosclerotic plaques. However, IVUS cannot be used for routine evaluation of plaque characteristics because of its invasiveness and the related increased risk during the procedure, additional time required, and cost [3].

∗ Corresponding author. Tel.: +86 29 87679770; fax: +86 29 87679775. ∗∗ Co-corresponding author. E-mail addresses: [email protected] (D. Gao), [email protected] (J. Yang). 1 These authors contributed equally to this work. 0021-9150/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2011.08.022

The development of high-speed spiral computed tomography allows coronary artery visualization and the detection of coronary stenoses. In addition it has been suggested as a novel, noninvasive modality for coronary atherosclerotic plaque detection, characterization, quantification and even coronary risk stratification [4,5]. Over the last decade, there has been a growing interest in assessment of coronary atherosclerotic plaque by using contrast-enhanced multi-detector spiral CT (MDCT). These trials investigated diagnostic accuracy, measurement of plaque volume, and vascular morphology, comparing MDCT with IVUS as the reference standard. Some studies have shown the ability of CT to detect coronary plaques. However, the ability of MDCT to detect and quantify coronary atherosclerotic plaque in vivo has never been systematically validated because of the small sample sizes. We therefore performed a meta-analysis to explore the potential diagnostic value of CT in detecting coronary plaques.

604

D. Gao et al. / Atherosclerosis 219 (2011) 603–609

1. Methods 1.1. Search strategy We searched for articles in MEDLINE via PubMed and EMBASE from 1966 through January 2011 using the terms “computed tomography”, “intravascular ultrasound” and “plaque”. When possible, we used the thesaurus terms of the specific database (MeSH terms for MEDLINE and EMTREE terms for EMBASE). We also reviewed the bibliographies of articles for possible reports. We included studies and abstracts that were presented at recent congresses (American College of Cardiology [2005–2010], American Heart Association [2005–2010], European Society of Cardiology [2005–2010]). 1.2. Study selection We included articles in any language that (1) compared coronary CT with IVUS as the reference standard; (2) had a prospective design; (3) included patients with suspected or known CAD; (4) involved multislice CT scanners (≥4 slice); and (4) clearly stated number of true positive, false positive, false negative, and true negative results for diagnosis of any plaque. When the report did not contain sufficient details to evaluate the validity of the study or outcome data were missing, we attempted to contact the authors by email. Two authors (NN and DG) reviewed all retrieved abstracts to identify potential reports, then retrieved the full text of potentially eligible articles. Disagreements were resolved by consensus, and if necessary, a third author (YG). 1.3. Data extraction and quality assessment Data were abstracted by use of specific data collection forms by two authors (DG and NN) and checked for accuracy (by YG). We extracted data on study population characteristics, relevant technical information about the CT and IVUS approach, and detailed reference standard specifications. We assessed methodological quality by the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool [6] and individually tailored the guidelines for scoring each item in our review as suggested [7].

Fig. 1. Methodological quality of included studies according to the Quality Assessment of Diagnostic Accuracy Studies tool.

We used the MIDAS and METANDI modules of STATA 11.0 (StataCorp, College Station, TX) for analysis. We used STATA and REVMAN 5.1 (Revman; The Cochrane Collaboration, Oxford, UK) to produce graphs. Heterogeneity of results across trials was assessed with a standard chi-square test with significance set at P < 0.10 and an I [2] statistic with significance set at I [2] >50%. A value of P < 0.01 was considered statistically significant. 2. Results The search revealed 483 reports; 386 reports were excluded after reading the abstract and a further 81 reports were excluded after reading the full report. We finally accepted 16 reports and added 1 report after searching congress abstracts [10]. Appendix Fig. 1 shows the selection of articles for the study. 2.1. Study characteristics

1.4. Statistical analysis For data analysis, we applied exact binomial rendition of the bivariate mixed-effects regression model developed by Van Houwelingen [8] for treatment-trial meta-analysis, modified for synthesis of diagnostic test data [9]. This model fits a twolevel model, with independent binomial distributions for the true positives and true negatives conditional on the sensitivity and specificity in each study and a bivariate normal model for the logit transforms of sensitivity and specificity between studies. On the basis of this model, we estimated mean logit sensitivity and specificity, with their standard errors and 95% confidence intervals (95% CIs); estimates of the between-study variability in logit sensitivity and specificity; and the covariance between them. We then back-transformed these quantities to the original receiveroperating characteristic (ROC) curve to obtain summary sensitivity, specificity, and likelihood ratios. We used the derived logit estimates of sensitivity and specificity and their respective variances to construct a hierarchical summary ROC curve, with summary operating points for sensitivity and specificity on the curves and a 95% confidence contour ellipsoid. To search for publication bias, we constructed effective sample size funnel plots versus the log diagnostic odds ratio and performed a regression test of asymmetry. To assess the agreement between investigators in evaluating methodological quality with the QUADAS tool, we calculated ␬ statistics.

The study characteristics are shown in Table 1. All of the reports described single-center prospective cohort studies. All of the reports used contrast-enhanced CT protocols to detect coronary plaques. The number of included patients ranged from 6 to 100. The include reports described the detection of plaques at the coronary segment level or 3 to 5 mm section levels. Seven studies [11–17] used a 16-slice CT scanner [8,10,18–24], a 64-slice CT scanner, 1 [25] a dual-source scanner, and 1 [26] involved 2 different types of scanners (64- and 320-slice CT). For most studies, CT evidence of plaque was defined as structures >1 mm2 within and/or adjacent to the coronary artery lumen. For all studies, IVUS evidence of plaque was defined as lesions of at least 0.5 mm located between the media and the intima. In total, 17 articles [10–26] described diagnostic accuracy of CT in detecting any coronary plaque, 8 [11,16–20,22,24] in detecting calcified plaques and 7 [11,17–20,22,24] in detecting none-calcified plaques. 2.2. Methodological quality Fig. 1 summarizes the methodological quality of the studies according the QUADAS tool. Table 1 shows how the studies scored on each item. We omitted items 6 and 7 of the original QUADAS tool (differential verification and incorporation bias) because IVUS was the only acceptable reference standard according to our inclu-

D. Gao et al. / Atherosclerosis 219 (2011) 603–609

605

Table 1 Characteristics of included studies. Study

Year

No. of patients

Mean age (SD) [Correct?]

Men

Disease status

Mean HR

Level of analysis

No. of segments

Caussin et al. Leber et al.

2004 2004

22 54

58 (5) 63 (7)

11 33

CAD CAD

NA 59(3)

Segment Section

42 525

13 41

Moselewski et al.

2004

26

62

17

58(9)

Site

100

Achenbach et al.

2004

22

58

14

59(6)

Segment

Leber et al.

2005

59

64 (10)

NA

<65

Leber et al.

2006

20

59 (9)

18

NA

Mieghem et al. Wu et al.

2006 2007

67 35

58 (11) 43–75

57 22

Wang et al.

2007

15

NA

NA

Ye et al. Iriart et al. Sun et al.

2007 2007 2008

12 20 26

68 (9.6) 53 (12) 56

17 17

Petranovic et al.

2009

11

NA

NA

Shen et al. Velzen et al.

2010 91 2010 106

64.8 (9.1) 57 (11)

53 64

Yu et al.

2010

17

67 (11)

12

Suspected CAD Suspected CAD Suspected CAD Suspected CAD CAD Suspected CAD Suspected CAD CAD ACS Suspected CAD Suspected CAD CAD Suspected CAD Suspected CAD

Kesarwani et al.

2010

30

NA

Suspected CAD

Segment model

IVUS type

CT typeslice

40 MHz 20 MHz

16 s 16 s

65

NA 3 mm Section Site

40 MHz

16 s

83

50

15

40 MHz

16 s

Segment

825

55

15

20 MHz

64 s

365

161

64 s

67 94

44 56

3 mm Section ROI 15

20 MHz

NA 58(6)

3 mm Section ROI Segment

30 MHz 40 MHz

16 s 16 s

NA

Segment

86

38

NA

NA

64 s

62(11) 57 NA

ROI Segment Segment

88 169 263

51 86 116

15 15 15

40 MHz 40 MHz 20 MHz

64 s 16 s 64 s

NA

Site and segment Segment Segment

122

60

17

40 MHz

64 s

91 528

67 329

15 17

40 MHz 20 MHz

NA

10 mm Section

114

37

NA

20 MHz

NA

Segment

146

136

15

NA

64 s 64 and 320 s Dualsource CT 64

NA NA

No. of plaques

CAD, coronary artery disease; ACS, acute coronary syndrome; ROI, region of interest; NA,unable to assess; IVUS, intravascular ultrasonography; CT, computed tomography.

sion criteria, and all of the included reports fulfilled these criteria. We added one item (Was the interobserver reproducibility of CT described? [interobserver reproducibility]) to the original QUADAS tool because interobserver reproducibility could also influence the study results. We excluded 1 report from quality evaluation because of lack of information. Our interrater reliability for assessing quality items was good ( = 0.81; P < 0.001). The studies were generally of moderate quality. Only 6 of the QUADAS items were met by more than 75% of the studies. All of the reports met the requirements for at least 4 of the items, but none fulfilled all of them.

2.4. Covariate analyses Table 2 shows the covariate analyses that reached statistical significance. For our comparison of CT scanner generations (4- to 16-slice scanners vs >16-slice scanners), we added the number of slices as a covariate to the bivariate model. The sensitivity and specificity were higher for scanners with >16 detector rows than for scanners with ≤16 rows (sensitivity 95% vs 83%, P < 0.001 and specificity 94% vs 92%, P < 0.001). In addition to scanner generation, blinded study design, quality of the study, and interobserver reproducibility also affected the results. 2.5. Probability of predictive value of CT for diagnosing coronary artery plaque

2.3. Diagnostic performance of CT in detecting plaque Among the 17 reports providing data on the diagnostic accuracy data of CT in detecting any coronary plaque, the pooled sensitivity was 92% (95% CI, 88–95%) and the pooled specificity 93% (95% CI, 90–96%) (Fig. 2a). For diagnosis of calcified plaque, the sensitivity and specificity were 93% (84–97%) and 98% (96–99%), respectively (Fig. 2b), and for diagnosis of noncalcified plaque (soft or fibrotic), the sensitivity and specificity dropped to 88% (81–93%) and 92% (89–95%), respectively (Fig. 2c). Fig. 3a shows the resulting hierarchical summary ROC curves for diagnosis of any plaque, with summary operating point for sensitivity and specificity on the cures and a 95% confidence and a 95% prediction contour around the point. For diagnosis of any plaque, the area under the ROC curve (AUC) was 0.98 (0.96–0.99) and for diagnosis of calcified and noncalcified plaque, the AUC was 0.99 (0.9–1.0) and 0.96 (0.94–0.97), respectively (Fig. 3b and c). We found no evidence of small-study effect for our meta-analysis (P = 0.56) (Appendix Fig. 2).

To determine the implications of our results for clinical use, we calculated the probability of CT in diagnosing any coronary plaque (posttest probability) with different prevalences (pretest probabilities), according to Bayes theorem (Fig. 4). CT had a good negative predictive value (0.84, 0.81–0.87) and positive predictive value (0.86, 0.83–0.89) over a wide range of pretest probabilities for diagnosing any plaque. Appendix Fig. 3 shows the likelihood ratio profile of CT (positive likelihood ratio <10; negative likelihood ratio <0.1) primarily as a tool for ruling out any plaque. Furthermore, a positive CT result increased the post-test probability markedly (43%, 82%, and 93% for pretest probabilities of 5%, 25%, and 50%, respectively) (Appendix Fig. 4) for segments with suspected plaques. 3. Discussion Alternative diagnostic approaches for the visualization of coronary artery plaques must prove their accuracy, reliability, and

606

D. Gao et al. / Atherosclerosis 219 (2011) 603–609

Fig. 2. Forest plot of eligible studies showing estimated sensitivities and specificities of included studies, as well as pooled values (♦), with corresponding 95% confidence intervals (95% CIs) (in brackets).

Fig. 3. Hierarchical summary receiver operating characteristic (HSROC) curve plots of sensitivity (SENS) and specificity (SPEC) for CT detecting any plaque, calcified and noncalcified. Mean sensitivity and specificity values (♦) are shown. AUC = area under the ROC curve.

D. Gao et al. / Atherosclerosis 219 (2011) 603–609

607

Table 2 Diagnostic accuracy data for CT in detecting coronary plaque. Group

Mean sensitivity (95% CI),%

Mean specificity (95% CI),%

AUC (95% CI)

Positive likelihood ratio(95% CI)

Negative likelihood ratio(95% CI)

Plaque (n = 17) Calcified plaque (n = 9) Non-calcified plaque (n = 7) Covariates Scanner generation ≤16 >16 P value Blinded study design Yes No or unclear P value Study quality QUADAS item > 8 QUADAS item≤ 8 P value Interobserver reproducibility 2 or more observers 1 observer P value

0.92 (0.88–0.95) 0.93 (0.84–0.97) 0.88 (0.81–0.93)

0.93 (0.90–0.96) 0.98 (0.96–0.99) 0.92 (0.89–0.95)

0.98 (0.96–0.99) 0.99 (0.98–1.00) 0.96(0.94–0.97)

13.9 (9.1–21.3) 44.1 (22.1–88.0) 11.7(7.5–18.2)

0.08 (0.05–0.14) 0.07 (0.03–0.17) 0.13 (0.08–0.21)

0.83 (0.76–0.91) 0.95 (0.92–0.97) P < 0.001

0.92 (0.89–0.98) 0.94 (0.91–0.97)

0.90 (0.85–0.95) 0.95 (0.91–0.98) P < 0.001

0.93 (0.90–0.97) 0.94 (0.89–0.98) P < 0.01

0.90 (0.85–0.95) 0.95 (0.91–0.98) P < 0.001

0.93 (0.90–0.97) 0.94 (0.89–0.98) P < 0.01

0.91 (0.85–0.96) 0.93 (0.89–0.97) P < 0.001

0.93 (0.89–0.98) 0.94 (0.90–0.97) P < 0.001

P < 0.001

95% CI, 95% confidence interval; QUADAS, Quality Assessment of Diagnostic Accuracy Studies; AUC, area under the receiver-operating characteristic curve.

prognostic value before they can be recommended for clinical practice. IVUS has the advantage of increased spatial resolution but is invasive and limited because of its low penetration depth. We performed a meta-analysis of the diagnostic performance of CT as a potential noninvasive method for detecting coronary artery plaque. We observed a good weighted sensitivity of 92% and specificity of 93% for CT in the detection of any coronary artery plaque as compared with IVUS. Moreover, the sensitivity and specificity of CT was affected by plaque composition: CT has higher diagnostic sensitivity and specificity for calcified plaques (93% and 98%) than noncalcified plaques (88% and 92%). Of note, we found significant heterogeneity in results among the included studies. Thus, our results should be interpreted cautiously, although our use of a random-effects model and a bivariate random-effects meta-analysis model should correct for this issue at least in part. Use of meta-regression may explain some of the sources of heterogeneity. The sensitivity for scanners with >16 detector rows was significantly higher than that for scanners with ≤16 rows (94% vs 85%, P < 0.001), which suggests that different generations of CT scanners may be a source of the heterogeneity as well.

Fig. 4. Probabilities of coronary artery plaques after multislice CT.

We found one systematic review of CT for detecting coronary artery plaques [27]. Our meta-analysis may be an update and extension of this review. The reviewers retrieved 9 articles [14–17,22,23,28–30] providing data on the diagnostic value of CT for coronary plaques (searched only through PubMed and articles published up to April 2008). We identified all of these articles as well and omitted 3 [28–30] because of lack of information. Because of the few reports investigated and small sample sizes of studies, the authors of the review did not provide pooled results for sensitivity and specificity. In comparison, our meta-analysis, with a bivariate approach, yielded pooled sensitivities and specificities for CT in diagnosing any coronary plaque of 92% and 93%. Our meta-analysis contains some limitations. First, not all reports explicitly stated whether the studies were prospective. To retain a reasonable number of articles for analysis, we excluded only those that clearly stated the studies were retrospective. However, including a prospective study design as a covariate to the bivariate statistical model (prospective design clearly stated vs unclear) did not significantly influence sensitivity or specificity data (data not shown). Second, our reports described the investigation of different sets of segments for comparison with IVUS. Most of the studies used the segment classification of the American Heart Association. Others used 3–10 mm sections or regions of interest. Few of the studies included all segments. Because some reports did not describe including all coronary arteries in the analysis and others gave no information on the exclusion of coronary arteries, comparability of our results is limited. Most of the reports described including only 1 or 2 vessels, suspected to be abnormal, which can lead to overestimation of the diagnostic accuracy of CT for detecting coronary artery plaques. Third, because meta-analysis combines or integrates the results of several independent studies, the quality and reliability depends on the quality of included studies. We used the QUADAS tool for assessing methodological quality of individual studies as they were reported. The included studies we reviewed scored poorly on blinding of reference standard and index test results, reporting of uninterruptible results, explaining withdrawals and interobserver reproducibility. The high number of items that scored “unclear” shows the need for further improvement in this regard. Fourth, the study concerns selected patient populations that are referred for high-risk patients who planed to undertook invasive coronary angiography or have CAD. These patients probably

608

D. Gao et al. / Atherosclerosis 219 (2011) 603–609

have more advanced disease, which is easier to detect than for instance the less advanced disease in asymptomatic subjects. This may greatly limits the generalizability of our findings. From a clinical perspective, CT may be very useful to rule out the presence of any plaques due to the high negative predictive value (84%). Nonetheless, the negative likelihood ratio appears of great interest since it allows a considerable reduction in the posttest probability and thus CT might help in risk reclassification from high to low or intermediate risk. Importantly, supporting data are emerging that patients without any evidence of atherosclerotic plaques on CT have excellent prognosis that is maintained over a relatively long period of time. In contrast, patients with atherosclerotic plaques on CT have been shown to have worse outcome [31,32]. On the basis of these, CT conceivable may have several advantages for clinical management and may allow improved risk stratification. The greatest disadvantage in use of CT to detect coronary artery plaque by now is the lack of randomized studies analyzing the effect of these tests at the patient or vessel level, which may allow earlier detection of disease, on patient management and outcomes. In this meta-analysis, we only provided pooled accuracy data of CT to detect coronary artery plaque at the segment level. Although we found posttest probabilities of CT for populations with different prevalences (pretest probabilities) of any plaques were good on the basis of our findings for the studies, large, well organized, randomized studies still needed to provide more information for CT in detecting coronary artery plaque at vessel or patient level before clinical use. With ongoing technical developments, CT might have more potential in detecting coronary plaques. Additional detector rows and improvements in temporal resolution will advance CT technology. The latest-generation 320-row CT scanners became commercially available in mid-2008 and have recently been installed [33]. We identified one study involving this equipment; the sensitivity and specificity of CT in detecting any plaque were both increased to 99%. Further improvements, including dualsource CT, increased gantry rotation speeds, adaptive multisegment reconstruction for improved temporal resolution, and high-pitch acquisitions, may also increase the diagnostic value of CT for detecting coronary plaques. In summary, CT should be considered the foremost noninvasive alternative to IVUS for detecting coronary artery plaques. Randomized studies at the patient level are needed to address the potential use of CT in triage to alter management and outcomes in patients with high risk, suspected or onset CAD. Acknowledgements This study was supported by a grant from National Natural Science Foundation of China (No. 30900617 to Dengfeng Gao No. 30970797 to Jian Yang), Research Fund for the Doctoral Program of Higher Education of China (No. 2008.6981036 to Dengfeng Gao) and the National Basic Research Program of China (973 Program, No. 2010CB7332603). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis.2011.08.022. References [1] Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statistics—2011 update: a report from the American Heart Association. Circulation 2011;123:e18–209. [2] Arnoldi E, Henzler T, Bastarrika G, Thilo C, Nikolaou K, Schoepf UJ. Evaluation of plaques and stenosis. Radiol Clin North Am 2010;48:729–44.

[3] Mueller C, Hodgson JM, Schindler C, Perruchoud AP, Roskamm H, Buettner HJ. Cost-effectiveness of intracoronary ultrasound for percutaneous coronary interventions. Am J Cardiol 2003;91:143–7. [4] van Werkhoven JM, Schuijf JD, Gaemperli O, et al. Incremental prognostic value of multi-slice computed tomography coronary angiography over coronary artery calcium scoring in patients with suspected coronary artery disease. Eur Heart J 2009;30:2622–9. [5] 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;46:552–7. [6] Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003;3:25. [7] Whiting PF, Weswood ME, Rutjes AW, Reitsma JB, Bossuyt PN, Kleijnen J. Evaluation of QUADAS, a tool for the quality assessment of diagnostic accuracy studies. BMC Med Res Methodol 2006;6:9. [8] van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21:589–624. [9] Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982–90. [10] Kesarwani M, Choi T, Honoris L, Shavelle DM, Budoff MJ. Evaluation of plaque morphology by 64-slice computed tomographic angiography compared to intravascular ultrasound in non-occlusive segments of coronary arteries. J Am Coll Cardiol 2010;55:A75.E704. [11] Wu WH, Lu B, Jiang SL, et al. Noninvasive detection and evaluation of coronary atheroselerotic plaques with multi-slice spiral CT: a comparative study with intravascular ultrasonograhy. Chin J Radiol 2007;41:1027–31. [12] Iriart X, Brunot S, Coste P, et al. Early characterization of atherosclerotic coronary plaques with multidetector computed tomography in patients with acute coronary syndrome: a comparative study with intravascular ultrasound. Eur Radiol 2007;17:2581–8. [13] Van Mieghem CA, McFadden EP, de Feyter PJ, et al. Noninvasive detection of subclinical coronary atherosclerosis coupled with assessment of changes in plaque characteristics using novel invasive imaging modalities: the Integrated Biomarker and Imaging Study (IBIS). J Am Coll Cardiol 2006;47: 1134–42. [14] Moselewski F, Ropers D, Pohle K, et al. Comparison of measurement of cross-sectional coronary atherosclerotic plaque and vessel areas by 16-slice multidetector computed tomography versus intravascular ultrasound. Am J Cardiol 2004;94:1294–7. [15] Leber AW, Knez A, Becker A, et al. Accuracy of multidetector spiral computed tomography in identifying and differentiating the composition of coronary atherosclerotic plaques: a comparative study with intracoronary ultrasound. J Am Coll Cardiol 2004;43:1241–7. [16] Caussin C, Ohanessian A, Ghostine S, et al. Characterization of vulnerable nonstenotic plaque with 16-slice computed tomography compared with intravascular ultrasound. Am J Cardiol 2004;94:99–104. [17] Achenbach S, Moselewski F, Ropers D, et al. Detection of calcified and noncalcified coronary atherosclerotic plaque by contrast-enhanced, submillimeter multidetector spiral computed tomography: a segment-based comparison with intravascular ultrasound. Circulation 2004;109:14–7. [18] Shen Y, Qian JY, Wang MH, et al. qualitative assessment of non-obstructive left main coronary artery plaques using 64-multislice computed tomography compared with intravascular ultrasound. Chin Med J 2010;123:827–33. [19] Petranovic M, Soni A, Bezzera H, et al. Assessment of nonstenotic coronary lesions by 64-slice multidetector computed tomography in comparison to intravascular ultrasound: evaluation of nonculprit coronary lesions. J Cardiovasc Comput Tomogr 2009;3:24–31. [20] Sun J, Zhang Z, Lu B, et al. Identification and quantification of coronary atherosclerotic plaques: a comparison of 64-MDCT and intravascular ultrasound. Am J Roentgenol 2008;190:748–54. [21] Ye HH, Kaneda H, Saito S, et al. Qualitative and quantitative evaluation of coronary plaques with 64-slice computed tomography in comparison with intravascular ultrasound. Zhonghua Xin Xue Guan Bing Za Zhi 2007;35: 648–51. [22] Leber AW, Becker A, Knez A, et al. Accuracy of 64-slice computed tomography to classify and quantify plaque volumes in the proximal coronary system: a comparative study using intravascular ultrasound. J Am Coll Cardiol 2006;47:672–7. [23] Leber AW, Knez A, Von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol 2005;46:147–54. [24] Wang YH, Zhang ZQ, Lu B. A preliminary investigated report of utilizing 64 multi-slice CT coronary angiography in the evaluation of coronary atherosclerotic plaques. Chin J Radiol 2007;41:1189–93. [25] Yu wei WZ, Zhang Jianjun, He Lang, Lang Lingfeng, Cheng Jianming, Feng Yue. Assessment of coronary plaque morphology and arterial remodeling by dualsource computed tomography: comparison with in-travascular ultrasound. Diagn Imaging Interv Radiol 2010;19:3. [26] van Velzen JE, Schuijf JD, de Graaf FR, et al. Diagnostic performance of noninvasive multidetector computed tomography coronary angiography to detect coronary artery disease using different endpoints: detection of significant stenosis vs detection of atherosclerosis. Eur Heart J 2010.

D. Gao et al. / Atherosclerosis 219 (2011) 603–609 [27] Springer I, Dewey M. Comparison of multislice computed tomography with intravascular ultrasound for detection and characterization of coronary artery plaques: a systematic review. Eur J Radiol 2009;71:275–82. [28] Komatsu S, Hirayama A, Omori Y, et al. Detection of coronary plaque by computed tomography with a novel plaque analysis system, ‘plaque map’, and comparison with intravascular ultrasound and angioscopy. Circ J 2005;69:72–7. [29] Carrascosa PM, Capunay CM, Garcia-Merletti P, Carrascosa J, Garcia MF. Characterization of coronary atherosclerotic plaques by multidetector computed tomography. Am J Cardiol 2006;97:598–602. [30] Caussin C, Ohanessian A, Lancelin B, et al. Coronary plaque burden detected by multislice computed tomography after acute myocardial infarction with near-normal coronary arteries by angiography. Am J Cardiol 2003;92:849–52. [31] van Werkhoven JM, Schuijf JD, Gaemperli O, et al. Prognostic value of multislice computed tomography and gated single-photon emission computed tomography in patients with suspected coronary artery disease. J Am Coll Cardiol 2009;53:623–32.

609

[32] Pundziute G, Schuijf JD, Jukema JW, et al. Prognostic value of multislice computed tomography coronary angiography in patients with known or suspected coronary artery disease. J Am Coll Cardiol 2007;49:62–70. [33] Dewey M, Zimmermann E, Deissenrieder F, et al. Noninvasive coronary angiography by 320-row computed tomography with lower radiation exposure and maintained diagnostic accuracy: comparison of results with cardiac catheterization in a head-to-head pilot investigation. Circulation 2009;120:867–75.