Diagnostic Performance of Positron Emission Tomography in the Detection of Coronary Artery Disease: A Meta-analysis 1

Transcription

1 Diagnostic Performance of Positron Emission Tomography in the Detection of Coronary Artery Disease: A Meta-analysis 1 Kiran R. Nandalur, MD, Ben A. Dwamena, MD, Asim F. Choudhri, MD Sirisha R. Nandalur, MD, Priya Reddy, Ruth C. Carlos, MD Rationale and Objectives. Although myocardial perfusion positron emission tomography (PET), using either cyclotronproduced ammonia or generator-produced rubidium 82, has reported excellent diagnostic capabilities in the detection of coronary artery disease (CAD) in individual studies, the technique is not widely used in practice. This may be driven by cost and availability or by unawareness of performance. The purpose of our study was to conduct an evidence-based evaluation of PET in the diagnosis of CAD. Materials and Methods. We examined studies from January 1977 to July 2007 using MEDLINE and EMBASE. A study was included if it (1) used PET as a diagnostic test for CAD and (2) used catheter x-ray angiography as the reference standard ( 50% diameter stenosis). Analysis was performed on a subject and coronary territory level. Results. Nineteen studies (1442 patients) met the inclusion criteria. On a patient level, PET demonstrated a sensitivity of 0.92 (95% confidence interval [CI]: ) and specificity of 0.85 (CI: ), with a positive likelihood ratio (LR ) of 6.2 (CI: ) and negative likelihood ratio (LR ) of 0.11 (CI: ). On a coronary territory level (n 1130), PET showed a sensitivity of 0.81 (CI: ) and specificity of 0.87 (CI: ), with an LR of 5.9 (CI: ) and an LR of 0.19 (CI: ). Conclusion. PET demonstrates excellent diagnostic properties in the diagnosis of CAD, especially at the patient level. The capabilities appear superior to those reported in meta-analyses for perfusion imaging with Tl-201 and sestamibi, or anatomical imaging with coronary MDCT angiography or MRA. Given that previous studies have found PET to be costeffective and the current findings of excellent sensitivity and specificity, the modality should be more widely considered as an initial test in the diagnosis of CAD. Key Words. Meta-analysis; positron emission tomography; coronary disease. AUR, 2008 Acad Radiol 2008; 15: From the Department of Radiology, University of Michigan Health System, Ann Arbor, MI (K.R.N., B.A.D., R.C.C.); Department of Nuclear, Veterans Affairs, Ann Arbor Health Care System, Ann Arbor, MI (B.A.D.); Department of Radiology, University of Virginia Health System, Charlottesville, VA (A.F.C.); and Department of Radiation Oncology, William Beaumont Hospital, Royal Oak, MI (S.R.N.). Current address: Department of Radiology, William Beaumont Hospital, 3601 W 13 Mile Rd, Royal Oak, MI (K.R.N., P.R.). Address correspondence to: K.R.N. nandalurk@yahoo.com AUR, 2008 doi: /j.acra

2 Academic Radiology, Vol 15, No 4, April 2008 META-ANALYSIS OF PET FOR CORONARY DISEASE Coronary artery disease (CAD), secondary to atherosclerosis, is the leading cause of mortality in the United States, underlying or contributing to approximately 650,000 deaths per year (1). While catheter-based x-ray angiography evaluation of the coronary lumen is generally used as the clinical reference standard, the invasive procedure carries small but significant risks; thus, a potential noninvasive alternative would be valuable. Radionuclide myocardial perfusion imaging is one of the most common used diagnostic tests in the assessment of CAD and largely performed by thallium-201 or a technetium-99m perfusion tracer via single-photon emission computed tomography (SPECT). However, myocardial perfusion positron emission tomography (PET), which has several advantages over SPECT, including better spatial resolution, higher counting efficiencies, and excellent attenuation correction, resulting in single-center studies demonstrating robust sensitivity and specificity, is less commonly performed throughout the world. This is likely driven by the greater availability and lesser costs of SPECT, but lack of awareness regarding the diagnostic properties of PET in the diagnosis of CAD may also be a factor. Moreover, PET has become more prevalent and less expensive in the last decade, given its widespread use in oncological imaging. The purpose of our study was to provide an evidence-based evaluation of the clinical utility of PET for coronary artery disease; thus, we performed a comprehensive meta-analysis of all currently published studies comparing PET with catheter-based x-ray angiography as the gold standard. METHODS Data Sources and Searches MEDLINE and EMBASE for English and non-english literature published from January 1977 to July 2007 were searched evaluating for the presence of CAD in native or non-native coronary arteries by PET and catheter-based x-ray angiography in the same patients. The search included medical subject headings for positron emission tomography, myocardial perfusion, and coronary angiography with the exploded term coronary artery disease. Moreover, we evaluated bibliographies of retrieved articles, review articles, and textbooks. The retrieved studies were examined for potentially duplicate or overlapping data. Given that meeting abstracts provide insufficient information regarding their data and lack of finality regarding the results, they were excluded. Study Selection A study was included if (a) it used PET as a diagnostic test for obstructive CAD, with 50% diameter stenosis selected as the threshold for significant CAD, using catheter-based x-ray angiography as the reference standard and (b) reported cased in absolute numbers of true positive (TP), false positive (FP), true negative (TN), and false negative (FN) results or stated data adequate to derive this information. Studies were eligible regardless of whether they were referred for suspected or known CAD and regardless of technique used for PET. Studies were excluded if performed in (a) phantom-only models, (b) animals, or (c) normal healthy volunteers without catheter-based x-ray angiography correlation or (d) included fewer than 10 patients. Data Extraction and Quality Assessment Data extraction was performed by two independent investigators. Inconsistencies were resolved by discussion and consensus. The following information was extracted from each study: first author, year of publication, journal, location of study, and study population (number of patients examined with both test and number of diagnostic and nondiagnostic); number of patients with CAD; and gender, mean age, and technical characteristics of PET. Data were recorded, as available, at the patient level and coronary territory level (left anterior descending, left circumflex, and right coronary artery). Study quality and applicability were assessed by a modified checklist based on the Quality Assessment Tool for Diagnostic Accuracy (QUADAS) guidelines by two independent investigators, with discrepancies solved by consensus (2). Data Synthesis and Statistical Analysis The main analysis was performed at the patient level, as most studies focused on this level of information. Secondary analyses were performed at the coronary territory level. We applied the bivariate mixed-effects regression model for treatment trial meta-analysis and modified for synthesis of diagnostic test data assuming binomial errors distribution for sensitivity and specificity (3). Betweenstudy variability was assessed assuming correlated normally distributed random effects for logit (sensitivity) and logit (specificity) with the degree of correlation between predictive of an implicit threshold effect. We derived summary sensitivity, and specificity as functions of the estimated model parameters. Likelihood ratios (LR) are metrics that express how much the odds change for the presence of CAD if there is a positive (abnormal) PET 445

3 NANDALUR ET AL Academic Radiology, Vol 15, No 4, April 2008 (positive likelihood ratio: LR ) and also how much the odds change for the presence of CAD with a negative PET (negative likelihood ratio: LR ). LR and LR are defined with the following formulas: LR sensitivity/ (1 sensitivity) and LR (1 sensitivity)/specificity. Assessment of Heterogeneity Heterogeneity of the results between the studies was assessed graphically by forest plots and statistically using the quantity I 2 that describes the percentage of total variation across studies attributable to heterogeneity rather than chance. I 2 can be calculated from basic results as I 2 100% (Q df)/q, where Q is Cochran s heterogeneity statistic and df is the degrees of freedom. I 2 lies between 0% and 100%. A value of 0% indicates no observed heterogeneity, and values greater than 50% may be considered substantial heterogeneity. Statistical analyses were performed with Stata 9.0 (Chicago, II). ruling out disease), PET is excellent at confirming and excluding CAD (23). Coronary Territory-Level Summary Performance Estimates Per-coronary territory meta-analysis of PET pooled 9 datasets with 1130 coronary territories and demonstrated a sensitivity of 0.81 (CI: ) and specificity of 0.87 (CI: ) (Fig. 2) with an LR of 5.9 (CI: ) and an LR of 0.19 (CI: ). The prevalence of CAD by territory was 43.4% (407 of 1130). Assessment of Heterogeneity Analysis at the patient level demonstrated no significant heterogeneity between PET studies in sensitivities (I , p.36) but significant heterogeneity involving specificities (I , p.001). At the coronary territory level, heterogeneity was present between study sensitivities (I , p.001) and specificities (I , p.03). Quality grading by study is shown in Table 3. RESULTS Database searches identified 25 potentially relevant citations. Nineteen studies were included, with 6 being excluded because (a) they had overlapping data or (b) it was not possible to calculate absolute figures from the presented data. Study and population characteristics of the included studies are summarized in Table 1 (4 22). Data on diagnostic accuracy were available for the 19 studies with a total of 1442 patients. Results of the individual studies on a per-patient and per-coronary territory level are summarized in Table 2. Patient-Level Summary Performance Estimates After pooling 14 datasets (840 patients after exclusion of 2 patients secondary to unsuccessful PET), PET demonstrated a sensitivity of 0.92 (95% confidence interval [CI]: ) and specificity of 0.85 (CI: ) (Fig. 1). The prevalence of CAD in this group was 77.4% (650 of 840). Overall, these summary estimates show excellent sensitivity and specificity for CAD at the patient level. Evaluating clinical utility, the LR for PET was 6.2 (CI: ), and the LR was 0.11 (CI: ). Using the rule of thumb that for a diagnostic test to be useful it should have a high LR ( 5) (i.e., good at ruling in a disease) and a low LR ( 0.2) (i.e., good at DISCUSSION PET has increased in availability and gained widespread acceptance for use in oncology diagnosis over the past decade, given the vast amount of evidence supporting its utility. However, although there is similar evidence regarding its usefulness in cardiac imaging, PET is not routinely used by many centers throughout the world. PET has been shown to be cost-effective in the work-up of CAD relative to exercise testing, SPECT, and immediate angiography (24). Moreover, myocardial perfusion imaging with PET has significant prognostic value for predicting cardiac events, such as death and myocardial infarction, including patients whose diagnosis is less certain after SPECT and obese patients (25). To date, a comprehensive analysis of the data analyzing the use of PET in the diagnosis of CAD has not been performed, which could potentially increase the awareness of the modality. Our study found in an overall patient population with a high prevalence of diseases, PET has excellent sensitivity (92%) and specificity (85%) for detecting luminal stenoses 50%. Moreover, the positive and negative likelihood conferred by an abnormal and normal test, respectively, is also exceptional. The ability of PET to detect CAD appears superior to other modalities that perform functional imaging of the 446

4 447 Table 1 Characteristics of Included Studies First Author (Ref.) Year Journal Patients (N) Excluded Male (%) Abramson (4) 2000 Journal of Nuclear Bateman (5) 2006 Journal of Nuclear Botsch (6) 1994 European Journal of Nuclear Go (7) 1990 Journal of Nuclear Gould (8) 1986 Journal of the American College of Grover-McKay (9) 1992 American Heart Journal Laubenbacher (10) 1993 Journal of Nuclear Marwick (11) 1992 Journal of the American Society of Echocardiography Muzik (12) 1998 Journal of the American College of Namdar (13) 2005 Journal of Nuclear Sampson (14) 2007 Journal of the American College of Santana (15) 2007 Journal of Nuclear Schelbert (16) 1982 American Journal of Simone (17) 1992 American Journal of Physiologic Imaging Stewart (18) 1991 American Journal of Mean Age, yr (SD) PET Radiotracer Stressor Selection Stenosis (10) F-18 deoxyglucose (FDG) Treadmill n 8, dipyridamole n 11 Women referred for chest pain 50% NS NS Rubidium-82 Dipyridamole Suspected CAD 50% 31 0 NS NS 27 rubidium-82/23 nitrogen-13 (N-13) (NS) Rubidium-82 Dipyridamole Matched casecontrol 70% 34 0 NS 56 (NS) Rubidium-82 Exercise Suspected CAD 50% Dipyridamolehandgrip stress Suspected CAD CFR (11) Rubidium-82 Dipyridamolehandgrip stress 16 with known CAD 50% 29 0 NS 61 (11) Nitrogen-13 (N-13) Dipyridamole or Suspected CAD 75% adenosine (4) Rubidium-82 Dipyridamolehandgrip Known CAD 50% stress (12) Nitrogen-13 (N-13) Adenosine 31 with CAD 70% (NS) Nitrogen-13 (N-13) PET/CT Known CAD 50% (15) Rubidium-82 PET/CT (dipyridamole, adenosine, dobutamine) Suspected CAD 70% 53 NS 29 NS Rubidium-82 PET/CT Known or 50% (adenosine) Suspected CAD NS Nitrogen-13 (N-13) Dipyridamole 32 with CAD/13 50% normal volunteers 225 NS (NS) Rubidium-82 Dipyridamole Suspected CAD 67% (12) Rubidium-82 Dipyridamolehandgrip stress Suspected CAD 50% Academic Radiology, Vol 15, No 4, April 2008 META-ANALYSIS OF PET FOR CORONARY DISEASE

5 NANDALUR ET AL Academic Radiology, Vol 15, No 4, April 2008 Table 1 (Continued) Mean Age, yr (SD) PET Radiotracer Stressor Selection Stenosis First Author (Ref.) Year Journal Patients (N) Excluded Male (%) NS 25 0 NS 53 (NS) Nitrogen-13 (N-13) Exercise 19 known CAD/6 normal volunteers Tamaki (19) 1985 European Journal of Nuclear 51 0 NS 56 (NS) Nitrogen-13 (N-13) Exercise Known CAD 50% Tamaki (20) 1988 Journal of Nuclear 287 NS NS NS Rubidium-82 Dipyridamole Suspected CAD 67 Williams (21) 1994 Journal of Nuclear (NS) Nitrogen-13 (N-13) Exercise 40 with known CAD 75% Yonekura (22) 1987 American Heart Journal NS, not specified. heart. Recently, the emerging technique of stress cardiac magnetic resonance imaging (MRI) using either perfusion imaging or wall-motion imaging has shown promise, with stress-induced wall motion abnormalities imaging demonstrating a sensitivity of 83% and specificity of 86% and perfusion imaging demonstrating a sensitivity of 91% and specificity of 81% (26). A recent meta-analysis demonstrated a sensitivity of 86% and specificity of 74% for SPECT for CAD (27), and a sensitivity and specificity of 84% and 82% for stress echocardiography for CAD (28). Although the numbers for stress MRI, SPECT, and echocardiography are good, PET seems to exhibit the best combination of sensitivity and specificity. Moreover, in head-to-head studies, PET has compared favorably in diagnostic capabilities relative to SPECT (5,7,18). The robust capabilities of PET in detection of CAD has several underlying explanations. PET has excellent spatial and temporal resolution relative to SPECT, with greater myocardial count density. Bateman et al. (5) found that the PET count exceeded that of SPECT by a factor of 2, despite shorter imaging times: 5 minutes for PET versus 16 minutes for SPECT. This better count density allows for superior quality images and reslicing at 3 mm rather than the conventional 6 mm. PET also has an estimated theoretical tomographic in-plane spatial resolution of 2 to 4 mm compared to 6 to 8 mm for Tc-99m SPECT (full width at half maximum), which permits superior discernment of the myocardium from adjacent structures. Next, the sensitivity of PET may be better secondary to the characteristics of the tracer. Relative to Tc- 99m SPECT perfusion tracers, Rb-82 chloride is well extracted by the myocardium and the uptake is more linearly related to increases in coronary blood flow, as opposed to Tc-99m tracers, which plateau at relatively low flows. Moreover, PET, as opposed to SPECT, shows promise in not only detecting luminal stenoses but also characterizing atherosclerotic plaque inflammation, which is important, as recent studies have demonstrated that plaque characteristics may play a key role in the development of symptoms (29 31). There are several limitations to our study. First, the prevalence of disease in the study populations was high (77.4%), likely secondary to selection/verification bias favoring patients with abnormal PET scans to undergo invasive angiography. Next, significant heterogeneity was identified in multiple performance characteristics; thus, the results and potential clinical application should be interpreted with caution. Next, the quality of the studies, many performed before 2000, are generally low (scale 448

6 Academic Radiology, Vol 15, No 4, April 2008 META-ANALYSIS OF PET FOR CORONARY DISEASE Table 2 Per Patient, Per Coronary Territory Analysis Analysis by Patient Analysis by Coronary Territory First Author (Ref) N TP (n) FN (n) FP (n) TN (n) Sensitivity Specificity N TP (n) FN (n) FP (n) TN (n) Sensitivity Specificity Abramson (4) NS NS NS NS NS NS NS Bateman (5) NS NS NS NS NS NS NS Botsch (6) NS NS NS NS NS NS NS Go (7) NS NS NS NS NS NS NS Gould (8) NS NS NS NS NS NS NS Grover-McKay (9) Laubenbacher (10) NS NS NS NS NS NS NS Marwick (11) NS NS NS NS NS NS NS Muzik (12) NS NS NS NS NS NS NS Namdar (13) NS NS NS NS NS NS NS Sampson (14) NS NS NS NS NS NS NS Santana (15) Schelbert (16) NS NS NS NS NS NS NS Simone (17) NS NS NS NS NS NS NS Stewart (18) NS NS NS NS NS NS NS Tamaki (19) NS NS NS NS NS NS NS Tamaki (20) Williams (21) NS NS NS NS NS NS NS Yonekura (22) NS, not specified. Figure 1. Forest plots of patient-level sensitivity and specificity. A, Forest plot of sensitivity of PET compared with coronary angiography (CXA). B, Forest plot of specificity of PET. Filled box indicates point estimate of each study (area indicates relative contribution of the study to meta-analysis); horizontal line, 95% confidence interval (95% CI). 1 10, mean score 5.8, standard deviation 1.7), with many not providing comprehensive data at the patient and coronary territory levels. Next, the bias to publish studies with positive results, publication bias, confounds evaluation. Finally, although the availability of PET is increasing, the modality, specifically the equipment and radiotracer, is expensive and not widely available outside the United States. Moreover, radiation exposure remains a concern, especially when performed in conjunction with CT. For example, in a recent study by Sampson et al. 449

7 NANDALUR ET AL Academic Radiology, Vol 15, No 4, April 2008 Figure 2. Forest plots of coronary territory level sensitivity and specificity. A, Forest plot of sensitivity of PET compared with coronary angiography (CXA). B, Forest plot of specificity of PET. Filled box indicates point estimate of each study (area indicates relative contribution of the study to meta-analysis); horizontal line, 95% confidence interval (95% CI). Table 3 Quality Assessment First Author (Ref) Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 Item 7 Item 8 Item 9 Item 10 Score Abramson (4) Yes Yes Yes No Yes Yes Yes Yes No No 7 Bateman (5) Yes No Yes No No Yes Yes Yes Yes Yes 7 Botsch (6) No No Yes No No Yes Yes No No No 3 Go (7) Yes No Yes No No Yes Yes No Yes No 5 Gould (8) Yes No Yes No Yes Yes Yes No No No 5 Grover-McKay (9) Yes No Yes No Yes Yes Yes No No No 5 Laubenbacher (10) No No Yes No Yes Yes Yes Yes No No 5 Marwick (11) Yes Yes Yes No Yes Yes Yes No Yes No 7 Muzik (12) Yes Yes No No Yes Yes Yes No No No 5 Namdar (13) Yes No Yes No Yes Yes Yes No No No 5 Sampson (14) Yes No Yes Yes Yes Yes Yes Yes Yes Yes 9 Santana (15) Yes No Yes No Yes Yes Yes Yes Yes Yes 8 Schelbert (16) No No Yes No Yes Yes Yes No Yes No 5 Simone (17) Yes No Yes No No Yes Yes Yes Yes No 6 Stewart (18) Yes No Yes No Yes Yes Yes No Yes Yes 7 Tamaki (19) No No Yes No No Yes No No No No 2 Tamaki (20) Yes Yes Yes No Yes Yes Yes No Yes No 7 Williams (21) Yes No Yes No No Yes Yes No Yes No 5 Yonekura (22) Yes No Yes No Yes Yes Yes No Yes Yes 7 Item 1: Was the population clinically relevant, defined as a group of patients covering the spectrum of disease that is likely to be encountered in the current or future use of the test? Item 2: Was there complete verification by the reference standard? Item 3: Was there blinded interpretation of the test results? Item 4: Was there consecutive patient selection? Item 5: Was there prospective enrollment of patients? Item 6: Was there adequate description and quality of the imaging procedure? Item 7: Was the quality of the reference test technically adequate? Item 8: Was there adequate clinical description of the patient population? Item 9: Was the sample size 35 patients? Item 10: Was there adequate reporting of results, including summary and subgroup indices of accuracy? 450

8 Academic Radiology, Vol 15, No 4, April 2008 META-ANALYSIS OF PET FOR CORONARY DISEASE (14) using PET CT, the effective radiation dosimetry was 13.8 msv (12 msv from the rest-stress Rb-82 injections, 1.4 msv from the two CT transmission scans, and 0.4 from the two scout scans), which is relatively high. In conclusion, an evidence-based evaluation of PET demonstrates excellent sensitivity and specificity in diagnosis of coronary artery disease. Based on the current findings, increasing availability, and cost-effectiveness, PET should have an expanding role in the realm of noninvasive cardiac imaging. REFERENCES 1. Thom T, Haase N, Rosamond W, et al., for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics: 2006 Update: A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2006; 113:e85 e Whiting P, Rutjes AW, Reitsma JB, et al. 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: Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: Abramson BL, Ruddy TD, dekemp RA, et al. Stress perfusion/metabolism imaging: A pilot study for a potential new approach to the diagnosis of coronary disease in women. J Nucl Cardiol 2000; 7: Bateman TM, Heller GV, McGhie AI, et al. Diagnostic accuracy of rest/ stress ECG-gated Rb-82 myocardial perfusion PET: Comparison with ECG-gated Tc-99m sestamibi SPECT. J Nucl Cardiol 2006; 13: Botsch H, Beringer K, Petersen J, et al. Single-photon emission tomography studies of rubidium-81 in the detection of ischaemic heart disease, using a stress-reinjection protocol. Eur J Nucl Med 1994; 21: Go RT, Marwick TH, MacIntyre WJ, et al. A prospective comparison of rubidium-82 PET and thallium-201 SPECT myocardial perfusion imaging utilizing a single dipyridamole stress in the diagnosis of coronary artery disease. J Nucl Med 1990; 31: Gould KL, Goldstein RA, Mullani NA, et al. Noninvasive assessment of coronary stenoses by myocardial perfusion imaging during pharmacologic coronary vasodilation. VIII. Clinical feasibility of positron cardiac imaging without a cyclotron using generator-produced rubidium-82. J Am Coll Cardiol 1986; 7: Grover-McKay M, Ratib O, Schwaiger M, et al. Detection of coronary artery disease with positron emission tomography and rubidium 82. Am Heart J 1992; 123: Laubenbacher C, Rothley J, Sitomer J, et al. An automated analysis program for the evaluation of cardiac PET studies: Initial results in the detection and localization of coronary artery disease using nitrogen-13- ammonia. J Nucl Med 1993; 34: Marwick TH, Nemec JJ, Stewart WJ, et al. Diagnosis of coronary artery disease using exercise echocardiography and positron emission tomography: Comparison and analysis of discrepant results. J Am Soc Echocardiogr 1992; 5: Muzik O, Duvernoy C, Beanlands RS, et al. Assessment of diagnostic performance of quantitative flow measurements in normal subjects and patients with angiographically documented coronary artery disease by means of nitrogen-13 ammonia and positron emission tomography. J Am Coll Cardiol 1998; 31: Namdar M, Hany TF, Koepfli P, et al. Integrated PET/CT for the assessment of coronary artery disease: A feasibility study. J Nucl Med 2005; 46: Sampson UK, Dorbala S, Limaye A, et al. Diagnostic accuracy of rubidium-82 myocardial perfusion imaging with hybrid positron emission tomography/computed tomography in the detection of coronary artery disease. J Am Coll Cardiol 2007; 49: Santana CA, Folks RD, Garcia EV, et al. Quantitative (82)Rb PET/CT: Development and validation of myocardial perfusion database. J Nucl Med 2007; 48: Schelbert HR, Wisenberg G, Phelps ME, et al. Noninvasive assessment of coronary stenoses by myocardial imaging during pharmacologic coronary vasodilation. VI. Detection of coronary artery disease in human beings with intravenous N-13 ammonia and positron computed tomography. Am J Cardiol 1982; 49: Simone GL, Mullani NA, Page DA, et al. Utilization statistics and diagnostic accuracy of a nonhospital-based positron emission tomography center for the detection of coronary artery disease using rubidium-82. Am J Physiol Imaging 1992; 7: Stewart RE, Schwaiger M, Molina E, et al. Comparison of rubidium-82 positron emission tomography and thallium-201 SPECT imaging for detection of coronary artery disease. Am J Cardiol 1991; 67: Tamaki N, Yonekura Y, Senda M, et al. Myocardial positron computed tomography with 13N-ammonia at rest and during exercise. Eur J Nucl Med 1985; 11: Tamaki N, Yonekura Y, Senda M, et al. Value and limitation of stress thallium-201 single photon emission computed tomography: Comparison with nitrogen-13 ammonia positron tomography. J Nucl Med 1988; 29: Williams BR, Mullani NA, Jansen DE, et al. A retrospective study of the diagnostic accuracy of a community hospital-based PET center for the detection of coronary artery disease using rubidium-82. J Nucl Med 1994; 35: Yonekura Y, Tamaki N, Senda M, et al. Detection of coronary artery disease with 13N-ammonia and high-resolution positron-emission computed tomography. Am Heart J 1987; 113: Sistrom CL, Garvan CW. Proportions, odds, and risk. Radiology 2004; 230: Patterson RE, Eisner RL, Horowitz SF. Comparison of cost-effectiveness and utility of exercise ECG, single photon emission computed tomography, positron emission tomography, and coronary angiography for diagnosis of coronary artery disease. Circulation 1995; 91: Yoshinaga K, Chow BJ, Williams K, et al. What is the prognostic value of myocardial perfusion imaging using rubidium-82 positron emission tomography? J Am Coll Cardiol 2006; 48: Nandalur KR, Dwamena BA, Choudhri AF, Nandalur MR, Carlos RC. Diagnostic performance of stress cardiac magnetic resonance imaging in the detection of coronary artery disease: A meta-analysis. J Am Coll Cardiol 2007; 50: Underwood SR, Anagnostopoulos C, Cerqueira M, et al. Myocardial perfusion scintigraphy: The evidence. Eur J Nucl Med Mol Imaging 2004; 31: Bax JJ, Van der Wall EE, De Roos A, et al. State of the art and future directions. In: Zaret Bl, Beller GA, eds. Clinical Nuclear. Philadelphia: Mosby, Nandalur KR, Baskurt E, Hagspiel KD, et al. Calcified carotid atherosclerotic plaque is associated less with ischemic symptoms than is noncalcified plaque on MDCT. AJR Am J Roentgenol 2005; 184: Nandalur KR, Hardie AD, Raghavan P, et al. Composition of the stable carotid plaque: Insights from a multidetector computed tomography study of plaque volume. Stroke 2007; 38: Tawakol A, Migrino RQ, Bashian GG, et al. In vivo 18F-fluorodeoxyglucose positron emission tomography imaging provides a noninvasive measure of carotid plaque inflammation in patients. J Am Coll Cardiol 2006; 48: