Noninvasive assessment of coronary artery bypass grafts with 64-slice computed tomographic angiography

, , , , , ,
Cardiology Review® Online, January 2008, Volume 26, Issue 1

We evaluated the accuracy of the 64-slice computed tomography (CT) angiography scanner in subjects who were scheduled to undergo invasive angiography for possible stenosis in coronary artery bypass grafts. Results showed that the improved resolution of CT scanners with 64-slice technology allowed for a precise delineation of bypass graft occlusion or stenosis. It permitted an accurate noninvasive assessment of bypass grafts, even in subgroups of subjects with suboptimal scan conditions, such as those with arrhythmias or higher heart rates.

The recurrence of angina over the long term has been reported in up to 30% of patients who have undergone coronary artery bypass graft (CABG) surgery.1 Within 5 years after surgery, 25% of CABGs develop occlusive disease.2

The technology of multislice computed tomography (CT) has steadily improved in recent years and has increasingly made accurate noninvasive CABG imaging possible.3,4 Patients with arrhythmias have not been included in studies assessing the reliability of 4- and 16-slice CT in identifying the patency of bypass grafts. These studies, therefore, do not reflect actual conditions and restrict the usefulness of the technique to patients without arrhythmias.4 With the recent introduction of the 64-slice CT angiography scanner, which has enhanced spatial and temporal resolution, we conducted a prospective study to determine the system's ability to correctly reveal significant bypass graft disease in symptomatic patients scheduled for invasive angiography, regardless of the presence of arrhythmias.

Subjects and methods

We evaluated 138 consecutive patients who had previously undergone CABG surgery using the 64-slice CT angiography scanner about 1 day before they were scheduled to undergo invasive angiography for possible bypass graft occlusion. Subjects with arrhythmias were included in the study; those with known iodinated contrast agent allergy or with a serum creatinine level > 1.8 mg/dL were excluded.

Details regarding the preparation of the patients and the 64-slice CT angiography procedure have been previously published.5,6 The complete bypass graft was scanned. The amount of contrast medium was modified based on the contrast dye flow rate and the length of the scan. All bypass grafts were assessed by 2 investigators who were blinded to the angiographic outcomes but had knowledge of the surgical bypass graft report. We excluded all bypass grafts that had been treated with placed stents. Grafts with ≥ 50% luminal narrowing at any location were considered diseased.


Computed tomography angiography results were compared with traditional invasive angiography, which was carried out using standard procedures. Invasive angiograms were assessed by 2 investigators who were blinded to the 64-slice CT angiography results for the occurrence of luminal narrowing ≥ 50% in the grafts. The 64-slice CT angiography results were evaluated for positive and negative predictive value, sensitivity, and specificity. Subgroup analyses were performed for subjects with increased heart rates and arrhythmias. A value of ≤ .05 was accepted as statistical significance.


A total of 418 bypass grafts were evaluated over a 2-year period. Eighty-eight percent of subjects were men, and the mean age was 68 years (± 10 years). During the 64-slice CT scan, 36% of patients had a heart rate ≥ 65 beats/minute, and 30% of patients had arrhythmias. Twelve of the 418 bypass grafts with previously placed stents were excluded from the analysis.

Image quality was sufficient for analysis in 98% (n = 397) of the grafts. Assessment of 64-slice CT scans was not possible in 1 graft because of metallic clips adjacent to the graft course and in 8 grafts because of severe motion artifacts. Invasive angiography showed that 29% (n = 116) of the bypass grafts either had significant luminal narrowing (n = 32; 8%) or were totally occluded (n = 84; 21%), and 71% (n = 281) of the grafts were patent. The 64-slice CT angiography scans accurately identified all 84 occlusions and 113 of the 116 graft occlusions or stenoses (Table). The 64-slice CT scanner was not able to identify 2 short membranous-like stenoses close to the distal anastomosis site and 1 stenosis in the proximal third of a left internal mammary artery graft. The Table shows the results regarding diagnostic accuracy.

Table. Diagnostic accuracy of 64-slice computed tomography angiography in the detection of significant bypass graft disease.


% (n)


% (n)


% (n)


% (n)


% (n)

All bypass grafts

98 (397/406)

97 (113/116)

97 (273/281)

93 (113/121)

99 (273/276)

Without arrhythmias

100 (273/273)

98 (79/81)

97 (187/192)

94 (79/84)

99 (187/189)

With arrhythmias

95* (124/133)

97 (34/35)

97 (86/89)

92 (34/37)

99 (86/87)

Heart rate < 65 bpm

100 (262/263)

99 (68/69)

97 (187/193)

92 (68/74)

99 (187/188)

Heart rate ≥ 65 bpm

94* (135/143)

96 (45/47)

98 (86/88)

96 (45/47)

98 (86/88)


J Am Coll Cardiol.

PPV indicates positive predictive value; NPV, negative predictive value; bpm, beats per minute. * < .05. (Reprinted with permission from Meyer TS, Marinoff S, Hadamitzky M, et al. Improved noninvasive assessment of coronary artery bypass grafts with 64-slice computed tomographic angiography in an unselected patient population. 2007;49[9]:946-950.)


Thirty percent of subjects (42 of 138) had arrhythmias during scanning in a total of 131 grafts. All grafts in subjects without arrhythmias were able to be assessed, whereas only 95% of grafts in subjects with arrhythmias were able to be assessed ( < .01). As shown in the Table, the diagnostic accuracy was similar in all evaluable grafts among subjects with and without arrhythmias.


Patients with a heart rate ≥ 65 beats/minute (143 grafts; 50 of 138 subjects) had fewer evaluable grafts than patients with a heart rate < 65 beats/minute (94% vs 100%, respectively; < .01). As shown in the Table, the diagnostic accuracy was similar in all evaluable grafts regardless of heart rate.


The results of our study showed that the 64-slice CT scanner successfully detected bypass graft stenoses and occlusions in subjects with angina after CABG surgery, whether they were in sinus rhythm or had arrhythmias. Our findings are in agreement with other studies that compared invasive angiography with 4- and 16-slice CT scanning for the evaluation of bypass graft patency.4 The subjects in these prior studies had normal sinus rhythm, which does not reflect true clinical conditions, in which a significant number of patients have tachycardias or arrhythmias. We included subjects with arrhythmias and higher heart rates in our study, not just so-called ideal subjects.

As far as we know, this was the largest study comparing invasive angiography with 64-slice CT scanning among patients who had undergone CABG surgery. Sensitivity and specificity of the 64-slice CT angiography scanner for the detection of significant stenosis in bypass grafts were each 97%, demonstrating the enhanced accuracy of the scanner. The improved temporal resolution of 64-slice technology allows for a consistently accurate assessment of bypass grafts, even among patients with increased heart rates or arrhythmias.

The assessment of grafts with small calibers has been hindered with 4- and 16-slice CT systems due to artifacts caused by metal clips. This is now less of an obstacle because of the improved spatial resolution of the 64-slice CT scanner. Data acquisition with previous scanner types took much longer because of slower volume coverage of 4- and 16-slice CT systems; therefore, the most proximal parts of internal mammary arteries were usually not visualized to ensure a manageable breath-hold period for the patient. Even though these proximal segments do not often have major stenosis, complete visualization of the entire course of internal mammary artery grafts is now possible in 1 inspiratory breath-hold with the 64-slice CT system.

Motion artifacts in CT angiographies of patients with arrhythmias or higher heart rates have always made CT angiography reading difficult or impossible, especially with former CT scanner generations. Although the evaluability of bypass grafts is still lower in these 2 patient subgroups, the results for sensitivity and specificity in the detection of significant graft stenosis indicate that 64-slice CT angiography is highly accurate, even in suboptimal imaging conditions.

From a clinical standpoint, the additional assessment of the native coronary arteries would be desirable in symptomatic patients after CABG surgery. Evaluating native coronary arteries is usually difficult because atherosclerotic disease is frequently advanced, with severe calcifications in diffusely narrowed arteries of small calibers. The 64-slice CT system is not able to evaluate all of these small arteries for significant stenosis (we found one third of the native arteries supplied with patent bypass grafts in this study non-evaluable), even with the improvement in temporal and spatial resolution. This may restrict the diagnostic usefulness of CT angiography in ruling out significant disease in the native coronary arteries of patients after CABG surgery.


The results of our study showed that CABGs can be evaluated for patency and the presence of significant stenosis using the 64-slice CT angiography system, even in patients with arrhythmias. This technique is useful as a routine noninvasive clinical test because of its high sensitivity and negative predictive value in unselected patients after CABG surgery.


We are indebted to the medical and technical staff members of the CT laboratories for their invaluable contribution, in particular, Christine Schwarzer, Brigitte Child, Gabriele Schmidbauer, and Christina Ardelean (all radiographers) for their expert technical assistance.