2004;77:800 4 MSCT OF CORONARY ARTERY BYPASS GRAFTS. Results. CABG With Adequate Diagnostic Quality

Transcription

1 Isotropic Half-Millimeter Angiography of Coronary Artery Bypass Grafts With 16-Slice Computed Tomography Marc Dewey, MD, Alexander Lembcke, MD, Christian Enzweiler, MD, Bernd Hamm, MD, and Patrik Rogalla, MD From the Department of Radiology, Charité, Humboldt-University Medical School, Berlin, Germany Background. Computed tomography (CT) with four detector rows and magnetic resonance imaging (MRI) are still of limited value for the assessment of coronary artery bypass grafts (CABG). We investigated the abilities of 16-slice CT in these patients. Methods. A retrospective analysis of all noninvasive coronary angiographies with multislice computed tomography (MSCT; Aquilion, Toshiba) on patients with CABG referred to our institution between October 2002 and April 2003 was conducted. MSCT angiography was performed using a standard protocol (0.5-seconds rotation time, mm detector collimation, 120 kv, 250 to 300 ma, and 0.25 pitch). None of the patients received -blockers to reduce the heart rate. Seventy-five CABGs (20 arterial grafts and 55 venous grafts) in 27 patients were evaluated for patency and adequate diagnostic quality by two radiologists in consensus. Results. All arterial and venous grafts were depicted with adequate diagnostic quality and were eligible for evaluation. Fifteen occlusions and five significant stenoses (at least 50%) could be identified. All of the proximal and 99% of the distal anastomoses were eligible. One distal anastomosis of an arterial graft was not assessable due to surgical clip artifacts. The length of the acquisition window was ms (range 71 to 234 ms). The majority of the patients (70%) had a heart rate above 65 beats/min. However, due to the improved temporal and spatial resolution none of the examinations had an insufficient image quality. Conclusions. MSCT angiography with 16 detector rows and an isotropic high resolution reliably depicts CABG with adequate diagnostic quality. (Ann Thorac Surg 2004;77:800 4) 2004 by The Society of Thoracic Surgeons Accepted for publication Aug 7, Address reprint requests to Dr Dewey, Department of Radiology, Charité, Medical School of the Humboldt University, Schumannstr. 20/21, Berlin, Germany; marc.dewey@charite.de. Magnetic resonance imaging (MRI) and computed tomography (CT) are promising methods for imaging coronary artery bypass grafts (CABG). MRI has a fair diagnostic accuracy regarding vein graft disease [1]. However, it is still limited by its insufficient resolution and limited coverage [2, 3], especially the distal anastomosis (connecting the CABG and the coronary artery), which is often not reliably assessable by MRI [4 6]. For more than two decades CT has been used to image CABGs [7]. Electron beam CT is a suitable test to detect occlusions of CABGs [8]. Furthermore, it was demonstrated that CT with four detector rows provides an adequate diagnostic quality for assessing the lumen of CABGs in about 60% to 80% of all patients [9, 10] and has a good accuracy in diagnosing stenoses [9, 11]. Nevertheless, the distal anastomoses continue to be hard to evaluate [8 10]. The introduction of multislice CT (MSCT) with 16 detector rows allows for isotropic imaging with a high resolution ( mm 3 ) and short acquisition time (shortest acquisition window per heartbeat is 62.5 ms). Because it is of utmost importance to develop a reliable and noninvasive modality for the follow-up of patients with CABGs, we analyzed the abilities of 16-slice CT. Patients and Methods All coronary angiographies performed with MSCT on patients with CABG referred to our institution between October 2002 and April 2003 were included in this retrospective analysis. The study was approved by the review board of the medical school. MSCT angiography was performed on a 16-slice CT scanner (Aquilion; Toshiba, Tustin, CA). Scan variables were as follows: 0.5-second rotation time, mm detector collimation, 0.35 to to 0.5 mm pixel size, 120 kv, 250 to 300 ma, 0.25 pitch, and breath hold time 40 seconds. None of the patients received -blockers before the examination to reduce the heart rate. A bolus of 120-mL iopromide (370 mg/ml) was intravenously injected into a cubital vein (3 ml/sec). The manual sure-start feature of the scanner was used to visualize the influx of the contrast medium. As soon as the contrast medium arrived in the pulmonary artery the patient was 2004 by The Society of Thoracic Surgeons /04/$30.00 Published by Elsevier Inc doi: /s (03)

2 Ann Thorac Surg DEWEY ET AL 2004;77:800 4 MSCT OF CORONARY ARTERY BYPASS GRAFTS Abbreviations and Acronyms CABG coronary artery bypass graft CT computed tomography GCV great cardiac vein LAD left anterior descending coronary artery LCX left circumflex coronary artery LIMA left internal mammary artery LV left ventricle MRI magnetic resonance imaging MSCT multislice computed tomography PDA posterior descending coronary artery PT pulmonary trunk RCA right coronary artery RVB right ventricular branch 801 CARDIOVASCULAR instructed to hold his breath. A simultaneous electrocardiogram (ECG) was recorded during the acquisition. The ECG was retrospectively used to assign source images to the respective phases of the cardiac cycle. A retrospective algorithm was applied to calculate axial images with a slice thickness of 0.5 mm using up to four different segments that are correlated to the raw data of up to four heartbeats ( Quadro segment reconstruction [12]). Images were reconstructed at intervals of 10% throughout the cardiac cycle. These axial images were evaluated on a Vitrea workstation (Vitrea 3.1; Vital Images, Plymouth, MN) and the cardiac phase with the least motion was chosen for further evaluation. Two radiologists in consensus evaluated the axial images, multiplanar reconstructions along and orthogonal to the graft, maximum intensity projections, and three-dimensional (3D) reconstructions in the optimal cardiac phase for the existence of significant stenoses or occlusions at the proximal and distal anastomosis and throughout the entire CABG. Image quality was graded insufficient or eligible (diagnostic quality without motion artifacts at least adequate for reliable assessment of the graft). Eligible grafts were further classified as nonstenotic, significantly stenotic (at least 50% luminal stenosis), or occluded. Fig 1. Three-dimensional representation of a normal CABG angiography. Patient with two venous CABGs (to the RCA and LCX) and an arterial graft (LIMA). Note the depiction of the distal anastomosis of the arterial graft (*) and the normally contrasted native LAD up to the apex of the heart. (CABG coronary artery bypass graft; LAD left anterior descending coronary artery; LCX left circumflex coronary artery; LIMA left internal mammary artery; RCA right coronary artery.) Results Twenty-seven patients were examined with 16-slice CT (3 female and 24 male patients; mean age 63 8 years; mean duration after CABG months). The patients underwent CABG angiography because of worsening angina (16 patients), or for routine (7 patients) and Table 1. Results of CABG Angiography With MSCT CABG Total CABG With Adequate Diagnostic Quality Anastomoses* With Adequate Diagnostic Quality Stenoses Occlusions Arterial grafts 20 (100%) 20 (100%) 19 (95%) 0 (0%) 3 (15%) Venous grafts 55 (100%) 55 (100%) 110 (100%) 5 (9%) 12 (23%) LAD 13 (100%) 13 (100%) 26 (100%) 2 (15%) 2 (15%) LCX 24 (100%) 24 (100%) 48 (100%) 1 (4%) 3 (13%) RCA 18 (100%) 18 (100%) 36 (100%) 2 (11%) 7 (39%) * The distal and proximal anastomoses of venous and the distal anastomoses of arterial grafts were evaluated. CABG coronary artery bypass graft; LAD left anterior descending coronary artery; LCX left circumflex coronary artery; MSCT multislice computed tomography; RCA right coronary artery.

3 802 DEWEY ET AL Ann Thorac Surg MSCT OF CORONARY ARTERY BYPASS GRAFTS 2004;77:800 4 Fig 2. Representative examples of venous CABGs including the distal anastomoses. The distal anastomoses (*) of three different patients are presented with (A) a maximum intensity projection, (B) a three-dimensional representation, and (C) a curved multiplanar reconstruction. Note the overlapping cardiac vein (arrowhead) in panel A, and the severe stenoses (arrows) in the native RCA and LAD in panels B and C. (CABG coronary artery bypass graft; GCV great cardiac vein; LAD left anterior descending coronary artery; LCX left circumflex coronary artery; LV left ventricle; PDA posterior descending artery; RCA right coronary artery.) postoperative control (4 patients). The 27 patients had a total number of 75 CABGs (mean: ) with 20 being arterial (left internal mammary artery) and 55 venous (Table 1). Diagnostic quality was adequate for all examinations (Figs 1 and 2). All 20 arterial and all 55 venous grafts were eligible for evaluation. All of the proximal (55) and 99% (74 of 75) of the distal anastomoses were eligible (Table 1 and Fig 2). One distal anastomosis of an arterial graft had to be excluded because of severe surgical clip artifacts. Six patients had relevant ancillary findings on MSCT: severe calcification of the aortic valve, metastases (from hitherto undetected prostate cancer), left ventricular aneurysm, left ventricular thrombus, paravalvular recess of the aortic valve, and aortic valve endocarditis. In the patient with prostate cancer one occlusion and a stenosis of a venous graft were detected, which could not be followed by conventional angiography. The left ventricular thrombus was successfully treated after diagnosis. MSCT demonstrated 15 occlusions (Table 1 and Fig 3) and five stenoses (all noncalcified; Table 1 and Fig 4). Conventional angiography confirmed nine occlusions, although exercise echocardiography confirmed ischemia in the area of two occlusions. One patient with two occlusions refused further workup, and in another patient (immediately after CABG) without symptoms the surgeons did not recommend conventional angiography. Two significant stenoses were confirmed by conventional angiography, although 1 patient was not followed because the referring cardiologist refused to do so as the patient had no symptoms. One stenosis, diagnosed as significant on MSCT, turned out to be nonsignificant on

4 Ann Thorac Surg DEWEY ET AL 2004;77:800 4 MSCT OF CORONARY ARTERY BYPASS GRAFTS 803 conventional angiography. In 3 patients without abnormal findings on MSCT angiography (five venous and two arterial grafts) were confirmed by conventional angiography. Seventy percent of the patients (19 of 27) had a heart rate above 65 beats/min (mean: beats/min, range 44 to 142 beats/min). Correct ECG gating was possible in all patients, no inadequate holding of breath occurred and none of the examinations had an insufficient image quality. The length of the acquisition window was ms (range: 71 to 234 ms). CARDIOVASCULAR Fig 3. Different representations of CABG occlusions in 2 patients: (A) axial image through an occlusion of a venous graft with the typical calcium deposit (arrow); (B) three-dimensional reconstruction of an arterial graft occlusion (arrows). Note the surgical clips along the former path of the arterial graft (arrows) in panel B. Despite the occlusion of the arterial graft in this patient, the LAD is normally contrasted because this native coronary artery had a nonsignificant stenosis. Note the distal anastomosis of the venous CABG to the RCA (*) and the right ventricular branch of the RCA. (CABG coronary artery bypass graft; LAD left anterior descending coronary artery; PT pulmonary trunk; RCA right coronary artery; RVB right ventricular branch.) Comment Noninvasive imaging of CABGs is one of the major challenges for cardiovascular imaging. At present neither CT nor MRI is capable of assessing CABGs reliably [6, 9, 10]. In our retrospective analysis of CABG angiography using a 16-slice CT scanner with isotropic half-millimeter resolution, all examinations had an adequate diagnostic quality for assessment. Only 1 of 75 distal anastomoses was not eligible for analysis. Occlusions were clearly depicted using this isotropic high-resolution approach. As only five stenoses were present on CT, the value of 16-slice CT in this regard remains to be determined. A considerable number of relevant ancillary findings were detected by MSCT. To achieve adequate diagnostic quality in 80% of the CABG examinations with 4-slice CT, it is essential to reduce the heart rate to less than 65 to 70 beats/min [10, 11]. In contrast, MSCT in our study yielded images of adequate quality in all examinations although the majority of the patients had a heart rate above 65 beats/min. Using MSCT, it is no longer necessary to reduce the heart rate by premedication. This improvement is mainly attributable to the enhanced temporal and spatial resolution of 16-slice CT. A major limitation of our study is the fact that invasive angiography is not available for comparison in all patients. However, because of the high resolution of this new MSCT method and the larger size of CABGs compared with native coronary arteries, we decided against a comparative conventional angiography that would have doubled the patients radiation exposure. This approach to the assessment of 16-slice CT is of course not adequate if one wants to compare the different diagnostic modalities in imaging of the much smaller native coronary arteries. Such a study would have to be done in direct comparison with the gold standard invasive coronary angiography. In addition, MRI, unlike CT, is useful for identifying grafts and recipient vessels with flow-limiting stenosis [13]. Nevertheless, the results presented here demonstrate the capability of 16-slice CT to reliably image CABGs. The higher percentage of examinations with an adequate diagnostic quality compared with 4-slice CT (62% [9], 80% [10]) is the major advantage of 16-slice CT. The results of this investigation demonstrate that multislice CT with 16 detector rows and isotropic halfmillimeter resolution reliably depicts CABGs with adequate diagnostic quality.

5 804 DEWEY ET AL Ann Thorac Surg MSCT OF CORONARY ARTERY BYPASS GRAFTS 2004;77:800 4 Fig 4. Curved multiplanar reconstruction along a venous coronary artery bypass graft demonstrating a significant stenosis (arrow). We thank Bettina Herwig for assistance in preparing the manuscript. References 1. Langerak SE, Vliegen HW, de Roos A, et al. Detection of vein graft disease using high-resolution magnetic resonance angiography. Circulation 2002;105: Kim WY, Danias PG, Stuber M, et al. Coronary magnetic resonance angiography for the detection of coronary stenoses. N Engl J Med 2001;345: Dewey M, Taupitz M. Coronary angiography by magnetic resonance imaging and computed tomography. Dtsch Med Wochenschr 2003;128: Vetter HO, Driever R, Mertens H, Kempkes U, Cramer BM. Contrast-enhanced magnetic resonance angiography of mammary artery grafts after minimally invasive coronary bypass surgery. Ann Thorac Surg 2001;71: von Smekal A, Knez A, Seelos KC, et al. A comparison of ultrafast computed tomography, magnetic resonance angiography and selective angiography for the detection of coronary bypass patency. Fortschr Rontgenstr 1997;166: Wittlinger T, Voigtlander T, Kreitner KF, Kalden P, Thelen M, Meyer J. Non-invasive magnetic resonance imaging of coronary bypass grafts. Comparison of the haste- and navigator techniques with conventional coronary angiography. Int J Cardiovasc Imaging 2002;18: Heuser L, Lackner K, Hauser H. Validity of computed tomography for the demonstration of patent or occluded aortocoronary vein grafts. A multicenter study. Fortschr Rontgenstr 1982;137: Jara FM, Kalush J, Kahn ML. Electron beam coronary angiography to assess patency in the off-pump coronary bypass graft. Ann Thorac Surg 2002;74:S Ropers D, Ulzheimer S, Wenkel E, et al. Investigation of aortocoronary artery bypass grafts by multislice spiral computed tomography with electrocardiographic-gated image reconstruction. Am J Cardiol 2001;88: Frohner S, Wagner M, Schmitt R, et al. Multi-slice CT of aortocoronary venous bypasses, and mammary artery bypasses: evaluation of bypasses and their anastomoses. Rontgenpraxis 2002;54: Treede H, Becker C, Reichenspurner H, et al. Multidetector computed tomography (MDCT) in coronary surgery: first experiences with a new tool for diagnosis of coronary artery disease. Ann Thorac Surg 2002;74:S Blobel J. New technologies for higher temporal and spatial resolution in cardiac multislice CTs. European Society of Cardiac Radiology (Ed.). Heidelberg: Springer, Langerak SE, Vliegen HW, Jukema JW, et al. Value of magnetic resonance imaging for the noninvasive detection of stenosis in coronary artery bypass grafts and recipient coronary arteries. Circulation 2003;107: