| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Department of Radiology (K.N., P.M.T.P., P.J.d.F.) and Thoraxcenter, Department of Cardiology (K.N., B.J.R., R.J.M.v.G.), Erasmus Medical Center, Dr Molewaterplein 40, Room D 220, Rotterdam 3015 GD, the Netherlands. Received July 21, 2002; revision requested September 26; final revision received March 24, 2003; accepted April 16. Address correspondence to K.N. (e-mail: koennieman@hotmail.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: ECG-gated contrast material–enhanced multi–detector row CT angiography was performed in 24 patients after bypass surgery. Two independent blinded observers evaluated all graft and coronary segments (
2.0-mm diameter) for occlusion and stenosis (50%–99% luminal reduction). Conventional angiography was regarded as the standard of reference. Descriptive parameters were calculated, and the results for arterial grafts, venous grafts, and coronary arteries, as well as for high and low heart rates, were compared by using a two-sided Fisher exact test.
RESULTS: The following results were obtained by observers 1 and 2, respectively: Of the 60 venous graft segments, 60 (100%) and 57 (95.0%) were assessable, with an overall detection of all 17 occlusions (both observers) and three (50.0%) and five (83.3%) of six stenoses. Of 26 arterial graft segments, 19 (73.1%) and 15 (57.7%) were assessable. In the assessable segments, four of four (100%) and two of three (66.7%) stenoses and occlusions were detected, while one and two obstructions were located in nonassessable segments. Of 211 coronary segments, 146 (69.2%) and 140 (66.4%) were assessable, and detection of 50%–100% obstruction yielded a sensitivity of 89.9% (71 of 79) and 79.4% (54 of 68) and a specificity of 74.6% (50 of 67) and 72.2% (52 of 72) for each observer. Unlike the assessment of venous and arterial grafts, assessment of the coronary arteries with multi–detector row CT was significantly better in patients with low heart rates (P < .01).
CONCLUSION: Multi–detector row CT allows noninvasive angiographic evaluation of both coronary arteries and bypass grafts in patients who have undergone bypass surgery. Multi–detector row CT is more effective in examining venous grafts compared with arterial grafts and diffusely diseased coronary arteries.
© RSNA, 2003
Index terms: Computed tomography (CT), angiography • Coronary angiography, 54.1244 • Coronary vessels, bypass graft, 54.455 • Coronary vessels, diseases, 54.76 • Coronary vessels, stenosis or obstruction, 54.76
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
All CT examinations were performed without complications. Each examination, including patient preparation and image reconstruction, generally required 10–30 minutes.
Image Processing and Data Analysis
A stack of approximately 250 overlapping transverse CT sections were further processed and analyzed on two separate workstations (O2 and Indigo 2; Silicon Graphics, Mountain View, Calif) with case-dependent application of post-processing techniques (Vitrea and VoxelView; Vital Images, Plymouth, Minn). Thin-slab maximum intensity projections allowed assessment of extended lengths of the vessels at once. In the presence of calcium or metal, the use of maximum intensity projection resulted in assessment-limiting overprojection, in which case double-oblique multiplanar reconstructions were then more suitable. Volume-rendered reconstructions were used for three-dimensional orientation and global presentation of results. Findings were then confirmed on the transverse CT source images.
All coronary artery segments (according to American Heart Association–American College of Cardiology guidelines) and bypass graft segments (consecutive graft anastomoses were regarded as separate segments) were independently evaluated by two investigators (K.N., P.M.T.P.) who were aware of the initial CABG procedure but were blinded toward the angiographic results (5). Each coronary or graft segment was classified as either interpretable or not interpretable according to the image quality. The venous graft segments were first screened for the presence of totally occluded segments. In the remaining segments, which were those that were considered patent by each observer, the image quality was then reevaluated, and interpretable segments were screened for the presence of stenotic lesions: luminal narrowing of 50%–99% of the luminal diameter. The arterial grafts were similarly assessed, but because of their relatively small diameter, no distinction was made between total occlusion and 50%–99% narrowing.
Only coronary segments with a minimal reference diameter of 2.0 mm were included for analysis; this assessment was made on the basis of conventional quantitative coronary angiographic findings. The presence or absence of calcium was noted for each coronary segment (K.N.). After evaluation of the interpretability of the multi–detector row CT data, all assessable segments were screened for the presence of substantial stenosis, including total occlusion (K.N., P.M.T.P.).
Conventional Coronary Angiography
Arterial catheterization and selective conventional angiography of the coronary arteries and bypass grafts were performed according to standard techniques. Quantitative coronary angiography, which involves catheter-derived image calibration and automated vessel contour detection, was performed of two orthogonal projections of the coronary arteries (CAAS software; Pie Medical, Maastricht, the Netherlands) to identify segments larger than 2.0 mm in diameter. An experienced interventional cardiologist (P.J.d.F.; 15 years experience) screened the angiographic images for stenotic lesions (50%–99% diameter reduction) or occlusions in the bypass grafts and coronary arteries. The diameter stenosis was determined by averaging the luminal narrowing from two orthogonal projections.
Statistical Analysis
The descriptive statistics were stratified for coronary artery segments, venous bypass graft segments, and arterial graft segments. Each graft, and each consecutive anastomosis in cases of sequential grafts, was regarded as a separate graft segment. Conventional quantitative coronary angiography was regarded as the standard of reference. The diagnostic parameters in patients with a low (<65 beats per minute) or high (65 beats per minute) average heart rate during the data acquisition were compared on the basis of the results of observer 1 (K.N.). Continuous variables were expressed as means and SDs. The diagnostic results from each of two observers in the detection of lesions in the assessable segments were expressed as sensitivity, specificity, negative predictive value, and positive predictive value. In addition, the overall sensitivity, which regards lesions in noninterpretable segments as false-negative assessments, was also calculated. Concordance between observers for the detection of obstructive lesions was calculated and expressed by using the 
value. Precision of the diagnostic parameters and interobserver variability were expressed by using a 95% CI. In addition, an overall evaluation of each complete arterial or venous graft was performed, in which the most severe lesion or the most proximally affected segment determined the accuracy of the observation. Results between arterial grafts, venous grafts, and coronary arteries were compared by using a two-sided Fisher exact test. The diagnostic performance of multi–detector row CT in depicting obstructive disease was compared between patients with high (<65 beats per minute) and low (65 beats per minute) heart rates; this comparison was performed with a two-sided Fisher exact test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
= 0.74 (95% CI: 0.52, 0.96). It was particularly good with respect to occlusions: = 0.91 (95% CI: 0.63, 1.00). In the evaluation of complete grafts, segments that were regarded as nonassessable for the detection of noncomplete stenosis were present in three of 23 venous grafts (for both observers). Total occlusion or noncomplete stenosis was correctly detected by observer 2 in 14 (93.3%) of 15 affected grafts. Observer 1 missed two stenoses, of which one was located in a segment regarded as nonassessable (sensitivity, 87.5%). A false-positive observation occurred once for each observer, and all other nondiseased grafts were correctly evaluated (specificity, 100%).
Arterial Bypass Grafts
Observer 1 assessed 19 (73.1%) of the 26 arterial graft segments and detected all four occlusions (100%) in the assessable segments; one noncomplete stenosis in a nonassessable segment was missed. Specificity, positive predictive value, and negative predictive value were 93.3% (14 of 15), 80.0% (four of five), and 100% (14 of 14), respectively (Table 4).
|
In the evaluation of complete grafts, observers 1 and 2 regarded one or more segments as unsuitable for analysis in five of 18 and in six of 18 arterial grafts, respectively. Observer 1 correctly detected the three grafts with occlusions (75.0%) but missed a stenosed graft that was considered nonassessable. Observer 2 detected one stenosed and one completely occluded arterial graft (50.0%) but missed two occluded grafts, of which one was regarded as nonassessable. The specificity was 100% (11 of 11) and 80.0% (8 of 10) for observers 1 and 2, respectively, if grafts with nonassessable sections were excluded.
Coronary Arteries
On the basis of findings at conventional angiography, an average of 8.8 ± 1.7 coronary segments with a minimal diameter of 2.0 mm were available per patient. Observers 1 and 2 found 146 (69.2%) and 140 (66.4%) of 211 segments interpretable, respectively. In the assessable segments, substantial obstruction, which included complete occlusions, was detected by observers 1 and 2, respectively, with a sensitivity of 89.9% (71 of 79) and 79.4% (54 of 68); specificity of 74.6% (50 of 67) and 72.2% (52 of 72); positive predictive value of 80.7% (71 of 88) and 73.0% (54 of 74); and negative predictive value of 86.2% (50 of 58) and 78.8% (52 of 66) (Table 5, Fig 3). The interobserver variability was reasonably good: = 0.68 (95% CI: 0.50, 0.86). Unfortunately, 23 (22.5%; observer 1) and 34 (33.3%; observer 2) of the 102 substantially obstructed segments were missed because of degraded image quality. Calcifications, which were the cause of many noninterpretable segments, were present in 142 (67.3%) of 211 coronary segments 2.0 mm or larger.
|
Influence of Heart Rate
Motion artifacts, caused by residual cardiac motion, were major contributors to the noninterpretability of multi–detector row CT images. In the group of patients with heart rates below 65 beats per minute, the coronary arteries were easier to interpret (P < .001) and showed a higher sensitivity (P < .01) compared with those in the group of patients with high heart rates. In the arterial grafts, there was a trend toward easier interpretability in the low heart rate group (P = .06). The venous grafts were easily assessable regardless of the heart rate (Table 6).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Even though the CT scanning range was not extended to cover the entire course of the proximal IMA grafts, in which no obstructions were found with conventional angiography, the CT scanning still required a long breath hold that could not be performed by a number of patients. Voluntary and cardiac motion were the major causes for nonassessability and misinterpretation. Extensive calcification of the coronary arteries and degenerated grafts, as well as the vascular clips in the proximity of arterial grafts (Fig 1), are the cause of beam-hardening and partial-volume artifacts that hinder assessment by suggesting or obscuring the obstruction.
Despite these technical limitations, our study results showed that multi–detector row CT angiography allows very accurate assessment of graft patency and, in addition, it provides relevant information concerning the presence of substantial obstructive disease in the bypass grafts and progression of disease of the coronary arteries. The diagnostic accuracy exceeds that of conventional and single–detector row helical CT (6,7). Evaluation of grafts with electron-beam CT was initially performed with individual images and later with three-dimensional reconstructions, which had good sensitivity (80%–100%) and specificity (82%–100%) for enabling detection of graft occlusion (8–12). In cases of adequate image quality, Achenbach et al (11) were able to detect 50%–99% stenosis in patent grafts as well (sensitivity, 84%). In addition, electron-beam CT flow studies have been performed to determine graft patency (13–15). Our study results confirm those of Ropers et al (16), who used multi–detector row CT to examine 182 grafts in 65 patients and reported a respective sensitivity and specificity of 97% and 98% in demonstrating graft occlusion and 75% and 92% in demonstrating substantial stenosis in grafts of adequate image quality (62%). In this study, however, the coronary arteries were not included in the evaluation.
Noninvasive follow-up of a patient who has undergone CABG cannot be restricted to visualization of the bypass grafts alone; it should include visualization of the coronary arteries. Results of a number of promising studies concerning the use of multi–detector row CT for noninvasive coronary angiography have been published. It appears that the diagnostic accuracy is reasonable, but complete assessment can be hindered by calcium deposits in the vessel wall and by motion artifacts, particularly in patients with high heart rates (17–20). Assessment of the native vessels in patients who have undergone bypass surgery is more challenging compared with that in patients who have an earlier stage of atherosclerotic disease. Advanced atherosclerotic degeneration results in small, diffusely narrowed vessels with an abundant presence of calcifications in the arterial wall, which complicates proper assessment of the vessel lumen. The high number of obstructions and the presence of a bypass graft anastomosis are obviously suggestive of the presence of a stenotic lesion in the proximal part of a coronary artery; the moderate assessability is only partly reflected in the diagnostic results.
Image degradation due to residual cardiac motion occurs predominantly in patients with higher heart rates. In the near future, faster rotation of the x-ray tube will increase the temporal resolution of the CT scanner. Until substantially faster CT scanners become available, motion artifacts can be reduced by administration of ß-receptor blocking medication prior to scanning in order to reduce the heart rate. The introduction of submillimeter–detector rows is expected to improve the assessment of severely calcified coronary segments and small distal branches.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, all authors; study concepts, K.N., P.J.d.F., R.J.M.v.G.; study design, K.N., R.J.M.v.G.; literature research, K.N.; clinical studies, K.N., R.J.M.v.G.; data acquisition, K.N., R.J.M.v.G.; data analysis/interpretation, K.N., P.M.T.P., B.J.R., P.J.d.F.; statistical analysis, K.N.; manuscript preparation and definition of intellectual content, K.N., P.M.T.P.; manuscript editing, K.N., P.M.T.P., B.J.R., P.J.d.F.; manuscript revision/review and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. A. Bluemke, S. Achenbach, M. Budoff, T. C. Gerber, B. Gersh, L. D. Hillis, W. G. Hundley, W. J. Manning, B. F. Printz, M. Stuber, et al. Noninvasive Coronary Artery Imaging: Magnetic Resonance Angiography and Multidetector Computed Tomography Angiography: A Scientific Statement From the American Heart Association Committee on Cardiovascular Imaging and Intervention of the Council on Cardiovascular Radiology and Intervention, and the Councils on Clinical Cardiology and Cardiovascular Disease in the Young Circulation, July 29, 2008; 118(5): 586 - 606. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Schroeder, S. Achenbach, F. Bengel, C. Burgstahler, F. Cademartiri, P. de Feyter, R. George, P. Kaufmann, A. F. Kopp, J. Knuuti, et al. Cardiac computed tomography: indications, applications, limitations, and training requirements: Report of a Writing Group deployed by the Working Group Nuclear Cardiology and Cardiac CT of the European Society of Cardiology and the European Council of Nuclear Cardiology Eur. Heart J., February 2, 2008; 29(4): 531 - 556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Simon, H. Baraki, J. Weidemann, W. Harringer, M. Galanski, and A. Haverich High-resolution 64-slice helical-computer-assisted-tomographical-angiography as a diagnostic tool before CABG surgery: the dawn of a new era? Eur. J. Cardiothorac. Surg., December 1, 2007; 32(6): 896 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Feuchtner, T. Schachner, J. Bonatti, G. J. Friedrich, P. Soegner, A. Klauser, and D. z. Nedden Diagnostic Performance of 64-Slice Computed Tomography in Evaluation of Coronary Artery Bypass Grafts Am. J. Roentgenol., September 1, 2007; 189(3): 574 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Malagutti, K. Nieman, W. B. Meijboom, C. A.G. van Mieghem, F. Pugliese, F. Cademartiri, N. R. Mollet, E. Boersma, P. P. de Jaegere, and P. J. de Feyter Use of 64-slice CT in symptomatic patients after coronary bypass surgery: evaluation of grafts and coronary arteries Eur. Heart J., August 1, 2007; 28(15): 1879 - 1885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M M Henneman, J D Schuijf, J W Jukema, H J Lamb, A de Roos, P Dibbets, M P Stokkel, E E van der Wall, and J J Bax Comprehensive cardiac assessment with multislice computed tomography: evaluation of left ventricular function and perfusion in addition to coronary anatomy in patients with previous myocardial infarction Heart, December 1, 2006; 92(12): 1779 - 1783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Ropers, F.-K. Pohle, A. Kuettner, T. Pflederer, K. Anders, W. G. Daniel, W. Bautz, U. Baum, and S. Achenbach Diagnostic Accuracy of Noninvasive Coronary Angiography in Patients After Bypass Surgery Using 64-Slice Spiral Computed Tomography With 330-ms Gantry Rotation Circulation, November 28, 2006; 114(22): 2334 - 2341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Achenbach Computed Tomography Coronary Angiography J. Am. Coll. Cardiol., November 21, 2006; 48(10): 1919 - 1928. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Budoff, S. Achenbach, R. S. Blumenthal, J. J. Carr, J. G. Goldin, P. Greenland, A. D. Guerci, J. A.C. Lima, D. J. Rader, G. D. Rubin, et al. Assessment of Coronary Artery Disease by Cardiac Computed Tomography: A Scientific Statement From the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology Circulation, October 17, 2006; 114(16): 1761 - 1791. [Full Text] [PDF]
|
|
|
|
|
|
C. M. Jones, T. Athanasiou, N. Dunne, J. Kirby, S. Attaran, A. Chow, S. Purkayastha, and A. Darzi Multi-slice computed tomography in coronary artery disease. Eur. J. Cardiothorac. Surg., September 1, 2006; 30(3): 443 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Feuchtner, W. Dichtl, T. Schachner, S. Muller, A. Mallouhi, G. J. Friedrich, and D. z. Nedden Diagnostic performance of MDCT for detecting aortic valve regurgitation. Am. J. Roentgenol., June 1, 2006; 186(6): 1676 - 1681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. U. Juergens, H. Seifarth, D. Maintz, M. Grude, M. Ozgun, T. Wichter, W. Heindel, and R. Fischbach MDCT Determination of Volume and Function of the Left Ventricle: Are Short-Axis Image Reformations Necessary? Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S371 - S378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Vernhet-Kovacsik, P. Battistella, R. Demaria, J. L. Pasquie, C. Bousquet, G. Dogas, F. Leclercq, B. Albat, and J. P. Senac Early Postoperative Assessment of Coronary Artery Bypass Graft Patency and Anatomy: Value of Contrast-Enhanced 16-MDCT with Retrospectively ECG-Gated Reconstructions Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S395 - S400. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Feuchtner, W. Dichtl, G. J. Friedrich, M. Frick, H. Alber, T. Schachner, J. Bonatti, A. Mallouhi, T. Frede, O. Pachinger, et al. Multislice Computed Tomography for Detection of Patients With Aortic Valve Stenosis and Quantification of Severity J. Am. Coll. Cardiol., April 4, 2006; 47(7): 1410 - 1417. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Pache, U. Saueressig, A. Frydrychowicz, D. Foell, N. Ghanem, E. Kotter, A. Geibel-Zehender, C. Bode, M. Langer, and T. Bley Initial experience with 64-slice cardiac CT: non-invasive visualization of coronary artery bypass grafts Eur. Heart J., April 2, 2006; 27(8): 976 - 980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yamamoto, F. Kimura, H. Niinami, Y. Suda, E. Ueno, and Y. Takeuchi Noninvasive Assessment of Off-Pump Coronary Artery Bypass Surgery by 16-Channel Multidetector-Row Computed Tomography. Ann. Thorac. Surg., March 1, 2006; 81(3): 820 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Jones, T. Athanasiou, P. P. Tekkis, V. Malinovski, S. Purkayastha, A. Haq, J. Kokotsakis, and A. Darzi Does Doppler echography have a diagnostic role in patency assessment of internal thoracic artery grafts? Eur. J. Cardiothorac. Surg., November 1, 2005; 28(5): 692 - 700. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Frazier, F. Qureshi, K. M. Read, R. C. Gilkeson, R. S. Poston, and C. S. White Coronary Artery Bypass Grafts: Assessment with Multidetector CT in the Early and Late Postoperative Settings RadioGraphics, July 1, 2005; 25(4): 881 - 896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. I. Stauder, M. Fenchel, H. Stauder, A. Kuttner, A. M. Scheule, U. Kramer, C. D. Claussen, and S. Miller Assessment of minimally invasive direct coronary artery bypass grafting of the left internal thoracic artery by means of magnetic resonance imaging J. Thorac. Cardiovasc. Surg., March 1, 2005; 129(3): 607 - 614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. R Mollet, F. Cademartiri, and P. J de Feyter Non-invasive multislice CT coronary imaging Heart, March 1, 2005; 91(3): 401 - 407. [Full Text] [PDF]
|
|
|
|
|
|
G. Aviram, R. Sharony, A. Kramer, N. Nesher, D. Loberman, Y. Ben-Gal, M. Graif, G. Uretzky, and R. Mohr Modification of Surgical Planning Based on Cardiac Multidetector Computed Tomography in Reoperative Heart Surgery Ann. Thorac. Surg., February 1, 2005; 79(2): 589 - 595. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Schuijf, J. J. Bax, J. W. Jukema, H. J. Lamb, H. W. Vliegen, E. E. van der Wall, and A. de Roos Noninvasive Evaluation of the Coronary Arteries With Multislice Computed Tomography in Hypertensive Patients Hypertension, February 1, 2005; 45(2): 227 - 232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Moreno and V. Fuster The year in atherothrombosis J. Am. Coll. Cardiol., December 7, 2004; 44(11): 2099 - 2110. [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Schuijf, J. J. Bax, J. W. Jukema, H. J. Lamb, H. W. Vliegen, L. P. Salm, A. de Roos, and E. E. van der Wall Noninvasive Angiography and Assessment of Left Ventricular Function Using Multislice Computed Tomography in Patients With Type 2 Diabetes Diabetes Care, December 1, 2004; 27(12): 2905 - 2910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Martuscelli, A. Romagnoli, A. D'Eliseo, M. Tomassini, C. Razzini, M. Sperandio, G. Simonetti, F. Romeo, and J.L. Mehta Evaluation of Venous and Arterial Conduit Patency by 16-Slice Spiral Computed Tomography Circulation, November 16, 2004; 110(20): 3234 - 3238. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
