Author + information
- Received May 24, 2006
- Revision received September 18, 2006
- Accepted October 9, 2006
- Published online March 6, 2007.
- Tanja S. Meyer, MD⁎,
- Stefan Martinoff, MD⁎,
- Martin Hadamitzky, MD†,
- Albrecht Will, MD⁎,
- Adnan Kastrati, MD†,
- Albert Schömig, MD† and
- Jörg Hausleiter, MD†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Jörg Hausleiter, Deutsches Herzzentrum München, Lazarettstrasse 36, 80636 München, Germany.
Objectives In this prospective evaluation we investigated the accuracy of 64-slice computed tomography angiography (CTA) in an unselected but symptomatic patient population for detection of stenoses in bypass grafts when compared with invasive angiography.
Background The assessment of significant stenosis in bypass grafts is important for patients with recurrent angina symptoms after bypass surgery.
Methods High-resolution 64-slice computed tomography (0.6 mm collimation, 330 ms gantry rotation time) and invasive angiography were performed in 138 consecutive patients with a total of 418 bypass grafts. Relevant stenosis was defined as diameter reduction ≥50%. During CTA, arrhythmias were present in 42 (30%) patients who were not excluded from the analysis.
Results The assessment of stenosis or occlusion of bypass grafts resulted in a sensitivity of 97%, specificity of 97%, and positive and negative predictive values of 93% and of 99%, respectively. The diagnostic accuracy for the detection of graft occlusion or stenosis did not differ between arterial and venous grafts. The evaluability of bypass grafts was significantly lower in patients with arrhythmias or with heart rates ≥65 beats/min during scanning. However, in the assessment of evaluable bypass grafts, no significant differences were detected in the diagnostic accuracy in these subgroups.
Conclusions This large prospective study demonstrates that 64-slice CTA is a reliable method for the assessment of bypass graft patency and stenoses even in an unselected “real-world” patient population.
Recurrence of angina is a frequent problem in the long-term follow-up of patients after coronary artery bypass grafting (CABG) (1). A total of 25% of grafts are found to be occluded within 5 years after surgery (2). Although arterial grafts usually remain free of progression of atherosclerotic disease, venous bypass grafts usually develop a stenotic and finally occlusive disease over time, which might result in myocardial ischemia.
Rapid advances in multislice computed tomography (MSCT) imaging technology have facilitated increasingly accurate noninvasive coronary artery and bypass graft imaging (3,4). However, previous studies investigating the accuracy of 4- and 16-slice computed tomography (CT) in the detection graft patency and the presence of graft stenosis have excluded patients with arrhythmias, which limits the usefulness of the method to ideal patient candidates without arrhythmias (4). A novel 64-slice MSCT scanner has recently been introduced that offers the potential for improved spatial and temporal resolution. Therefore, this large prospective evaluation assessed the accuracy of a 64-slice scanner in all patients previously scheduled to undergo invasive angiography for suspected bypass graft disease.
In 138 consecutive patients with a history of CABG and in whom invasive angiography was planned for suspected bypass graft disease, MSCT was performed usually 24 h before catheterization. Exclusion criteria were known contrast dye allergy or severe kidney failure (elevated serum creatinine >1.8 mg/dl). Patients with arrhythmias at the time of study inclusion or CT scanning were not excluded.
Patient preparation, MSCT angiography, and image interpretation
The methods for patient preparation and MSCT angiography have been described in detail elsewhere (5,6). In brief, contrast-enhanced computed tomography angiography (CTA) data (Sensation 64 Cardiac, Siemens Medical Solutions, Malvern, Pennsylvania) were acquired after vasodilation with nitroglycerin and administration of intravenous metoprolol in patients with a heart rate >60 beats/min. The scanning range included the entire course of venous grafts as well as the most proximal part of internal mammary artery (IMA) grafts at their subclavian origin, if these arterial grafts had been used for bypass surgery. The contrast dye volume (90 to 205 ml; 350 mg iodine/ml) was individually adapted to match the scan duration and the selected contrast dye flow rates. To minimize motion artifacts due to unwitting diaphragm movements, CT angiograms were acquired in the caudo-cranial direction, if the breath-holding period lasted ≥15 s. In patients with sinus rhythm, images were reconstructed in mid-diastole with a sharp kernel B36f, whereas in patients with arrhythmias the image reconstruction was performed usually in mid-diastole and end-systole. The methods for dose estimation of CTA have been described previously (5). Two investigators who were aware of the surgical CABG report but were blinded toward the angiographic results evaluated all bypass grafts with the use of axial slices and 3 thin-slab maximum intensity projections. Each graft was classified as either evaluable or not evaluable according to the image quality. Bypass grafts treated with placed stents were excluded from the analysis. The main analysis was performed on a per-graft basis, which considered a bypass graft diseased if there was a lumen narrowing ≥50% at any graft location. In the per-patient analysis, patients were classified as positive for significant graft disease if there was a significant stenosis in any bypass graft.
Conventional invasive angiography, which was the standard of reference for the comparison with MSCT results, was performed according to standard techniques. The angiograms were evaluated by 2 cardiologists blinded to the MSCT results. Quantitative coronary angiography was applied to determine lesion severity of diseased bypass grafts. The effective dose for invasive cardiac catheterization was estimated from the product of the measured dose-area product and a conversion coefficient k (k = 0.0022 mSv/[cGy × cm2] averaged between male and female models).
Results are expressed as counts (or proportions in percent) or as mean ± SD. The analysis was performed: 1) on a per-graft basis, evaluating the most severe lesion in a given bypass graft; and 2) on a per-patient basis, evaluating the presence of any significant bypass narrowing in a given patient. The diagnostic MSCT results in the detection of significant disease in the evaluable segments were expressed as sensitivity, specificity, and negative and positive predictive value with their respective 95% confidence interval (CI). In addition, a second per-patient–based analysis was performed on an “intention-to-diagnose” basis, in which bypass grafts determined as inconclusive by CTA were considered as significantly diseased by MSCT (lumen narrowing ≥50%).
Categorical variables were compared with chi-square analysis. Continuous variables were compared with the Student t test. Subgroup analyses focused on the diagnostic 64-slice CT performance in patients with arrhythmias, in patients with higher heart rates, and between arterial and venous bypass grafts. Statistical significance was accepted for p values < 0.05.
During the period from October 2004 through February 2006, 138 patients with a total of 418 bypass grafts were studied. Twelve bypass grafts with previously placed stents were excluded from the analysis. Table 1 summarizes the patient characteristics.
MSCT compared with invasive angiography for assessment of bypass grafts
Overall, 397 of 406 grafts (98%) demonstrated sufficient image quality for the assessment of bypass grafts by MSCT. In 9 grafts image quality was insufficient owing to motion artifacts (8 bypass grafts) or numerous metallic clips adjacent to the bypass graft (1 bypass graft). According to invasive angiography, 281 (71%) of these bypass grafts were patent and non-stenotic, whereas 116 (29%) presented with either complete occlusion (84 grafts, 21%) or significant stenosis (32 grafts, 8%). All 84 (100%) occlusions and 113 of 116 (97%) graft occlusions or stenoses were correctly identified by MSCT (Table 2,Figs. 1 and 2).⇓⇓ One stenosis in the proximal one-third of a left IMA graft and 2 very short membranous-like stenoses close to the distal anastomosis site in a venous and a radial artery graft were not identified with MSCT. The results for the diagnostic accuracy are summarized in Table 2. Of the 406 bypass grafts, arterial and venous bypass grafts were present in 147 (36.2%) and 259 (63.8%) grafts. The diagnostic accuracy did not differ between arterial and venous grafts (Table 2).
Influence of the presence of arrhythmias on MSCT accuracy for assessment of bypass grafts
Arrhythmias during scanning were present in 42 of 138 (30%) patients with a total of 131 grafts. The evaluability of bypass grafts in patients with arrhythmias was significantly lower than in patients without arrhythmias (95% vs. 100%, p < 0.01). No significant differences were detected for the diagnostic accuracy in evaluable grafts (Table 2).
Influence of heart rate on MSCT accuracy for assessment of bypass grafts
The evaluability of 143 bypass grafts in 50 of 138 (36%) patients with a mean heart rate of ≥65 beats/min was significantly lower than in patients with heart rates <65 beats/min (94% vs. 100%, p < 0.01). There were no significant differences in the diagnostic accuracy in the assessment of evaluable bypass grafts between patients with heart rates ≥65 beats/min and <65 beats/min (Table 2).
Diagnostic MSCT accuracy on a per-patient–based analysis
On a per-patient–based analysis, significant disease in bypass grafts could not be ruled out owing to a limited evaluability of the CT scan in 4 of 138 (3%) patients (motion artifacts in 3 patients, extensive metallic clips in 1 patient). In evaluable patients the sensitivity and specificity as well as positive predictive value (PPV) and negative predictive value (NPV) were 100% (95% CI 94% to 100%), 92% (95% CI 82% to 97%), 93% (95% CI 85% to 97%), and 100% (95% CI 93% to 100%), respectively. Including the non-evaluable patients for an “intention-to-diagnose”–based analysis, the resulting values for sensitivity and specificity as well as PPV and NPV were 100% (95% CI 94% to 100%), 87% (95% CI 76% to 93%), 89% (95% CI 79% to 94%), and 100% (95% CI 93% to 100%), respectively.
Comparison of radiation dose estimates and contrast dye volume
Radiation dose estimates and contrast dye volumes were compared between MSCT and invasive angiography for patients who underwent diagnostic invasive angiographies without subsequent percutaneous interventions. Radiation dose estimates for MSCT and invasive angiography were 17.8 ± 5.4 mSv and 8.8 ± 4.5 mSv, respectively (p < 0.05). Less contrast dye was administered with MSCT angiography (148 ± 22 ml vs. 223 ± 99 ml for MSCT vs. invasive angiography, p < 0.05).
This large prospective evaluation was carried out to assess the diagnostic accuracy of 64-slice CTA in the assessment of an unselected but symptomatic patient population after CABG. The study demonstrates that 64-slice CT consistently provides high-quality angiograms of bypass grafts that accurately delineate the presence of graft occlusion or stenosis irrespective of the presence of sinus rhythm or arrhythmias in these patients.
The current findings are consistent with and extend those of prior MSCT studies comparing 4- and 16-slice CT systems with invasive angiography for the assessment of patency and stenosis of bypass grafts (4). However, these studies were performed in a selected group of patients deemed ideal candidates for MSCT investigations; many “real world” patients (e.g., without sinus rhythm or with multiple ectopic beats, with heart rates >65 beats/min despite administration of beta blockers, or with pacemakers) were excluded. In contrast, 138 consecutive patients with a planned invasive angiography were included in the present study irrespective of the presence of the aforementioned exclusion criteria. To the best of our knowledge, this investigation represents the largest group of patients after CABG in which MSCT and invasive angiography are compared. The improved diagnostic efficacy of 64-slice CT is shown in a sensitivity of 97% and specificity of 97% for the detection of angiographically significant stenosis in bypass grafts. Although motion artifacts are already less prevalent in bypass grafts than in native coronary arteries, the improved temporal resolution of the studied 64-slice scanner allowed a reliable assessment of bypass grafts even in the presence of arrhythmias. Artifacts of metallic clips, which made particularly the assessment of arterial grafts more difficult with 4- and 16-slice CT systems, are less problematic with the improved spatial resolution of 64-slice CT. Furthermore, the most proximal part of IMA grafts were not visualized with previous CT systems, owing to slower volume coverage and the difficulties in maintaining a manageable breath-hold period. Although the presence of significant stenosis is less likely in these proximal segments, 64-slice CT systems now enable the complete visualization of the entire course of arterial grafts in 1 inspiratory breath-hold. The efficacy of this scanner in ameliorating imaging difficulties is also shown in the diagnostic accuracy to detect significant stenosis in the subgroups with less optimal imaging conditions, including patients with arterial grafts, which usually have a smaller caliber than venous grafts; patients with arrhythmias; or patients with higher heart rates. Although the evaluability of bypass grafts is lower, particularly in the later 2 subgroups, the derived results for sensitivity and specificity indicate that 64-slice CTA is highly accurate even in these patients with sub-optimal imaging conditions.
For a noninvasive study in a symptomatic patient after CABG, it would be desirable to include the assessment of native coronary arteries. The assessment of native coronary arteries in patients after CABG is challenging, owing to the advanced atherosclerotic disease with abundantly calcified and diffusely narrowed arteries with small dimensions. Despite the advances in spatial resolution with 64-slice CTA with an approximate voxel size of 0.5 mm, the resolution is still insufficient to accurately delineate significant stenosis in these small arteries. Therefore, these extensively diseased coronary segments are usually deemed nonevaluable, which might limit the diagnostic usefulness of CTA to rule out significant disease in clinical practice.
In summary, the newest technological development of 64-slice CTA permits the noninvasive assessment of venous and arterial bypass graft patency and stenoses with high diagnostic accuracy even in patients with arrhythmias during scanning. The high sensitivity and the excellent NPV demonstrate that this technology can be used in clinical routine as a noninvasive test for patients with suspected graft dysfunction.
The authors are indebted to the medical and technical staff members of the computed tomography laboratories for their invaluable contribution and in particular to Mrs. C. Schwarzer, B. Child, G. Schmidbauer, and C. Ardelean (all radiographers) for their expert technical assistance.
- Abbreviations and Acronyms
- coronary artery bypass grafting
- confidence interval
- computed tomography
- computed tomography angiography
- internal mammary artery
- multislice computed tomography
- negative predictive value
- positive predictive value
- Received May 24, 2006.
- Revision received September 18, 2006.
- Accepted October 9, 2006.
- American College of Cardiology Foundation
- Cameron A.A.,
- Davis K.B.,
- Rogers W.J.
- Fitzgibbon G.M.,
- Kafka H.P.,
- Leach A.J.,
- Keon W.J.,
- Hooper G.D.,
- Burton J.R.
- Achenbach S.,
- Ropers D.,
- Pohle F.K.,
- et al.
- Martuscelli E.,
- Romagnoli A.,
- D’Eliseo A.,
- et al.
- Hausleiter J.,
- Meyer T.,
- Hadamitzky M.,
- et al.
- Hausleiter J.,
- Meyer T.,
- Hadamitzky M.,
- Kastrati A.,
- Martinoff S.,
- Schömig A.