Coronary Angiography for the Evaluation of Coronary Artery Disease in Patients with Atrial Fibrillation: Initial Experience 1
Dilek Oncel, MD Guray Oncel, MD Ahmet Tastan, MD
Purpose: To prospectively evaluate the sensitivity and specificity of dual-source CT for significant coronary stenosis (⬎50%
narrowing) in patients with atrial fibrillation (AF), by using conventional coronary angiography as the reference stan- dard.
Materials and Methods:
Institutional Review Board approval and informed consent were obtained. Fifteen consecutive patients (nine men, six women; mean age, 58.47 years) were examined. Image quality (good, moderate, or poor) and significant stenosis (⬎50%) were evaluated by two radiologists blinded to the conventional coronary angiography results. Sensitivity, specificity, positive predictive value (PPV), and negative predictive values (NPV) were calculated. McNemar test was used to search for any significant difference between dual-source CT and conventional coronary angiography in helping detect coronary stenosis. statistics were calcu- lated for intermodality and interobserver agreement.
Results: Sixteen segments by reader 1 and 13 segments by reader 2 were considered as poor image quality and rejected for further analysis. All segments with good image quality were correctly diagnosed. The respective overall sensitiv- ity, specificity, PPV, and NPV values were 87%, 98%, 77%, and 99% for reader 1 and 80%, 99%, 80%, and 99%
for reader 2. No significant difference between dual-source CT and conventional coronary angiography was found in helping detect significant stenosis. statistics demon- strated good intermodality and excellent interobserver agreement.
Conclusion: Dual-source CT technology provides a temporal resolution that allows CT coronary angiography at higher heart rates and even with AF.
娀 RSNA, 2007
1From the Department of Radiology, Sifa Medical Center, Fevzipasa Boulevard No. 172/2, 35340 Basmane Izmir, Turkey. Received January 14, 2007; revision requested March 15; revision received March 23; accepted April 25;
final version accepted June 11. Address correspon- dence to D.O. (e-mail: [email protected]).
姝 RSNA, 2007
䡲 CARDIAC IMAGING
S
ince the introduction of multide- tector computed tomography (CT), one of the main limitations for cardiac imaging has been limited tem- poral resolution (1,2). In most standard protocols of CT angiography, including those with 64 –section multidetector CT scanners, -blocker administration is routinely used to reduce the heart rate, minimizing residual heart motion (3–8).Apart from high heart rates, ar- rhythmia remained as a contraindica- tion for CT angiography applications owing to between-beat variations, which led to inappropriate data sampling and resulted in severe motion artifacts (3–
8). With the newly introduced dual- source CT, the main improvement is the increase in temporal resolution. The system combines two arrays, each con- sisting of an x-ray tube and a 64 –sec- tion detector. With a rotation speed of 330 msec, a temporal resolution of 83 msec (one-fourth rotation) can be achieved independent of heart rate (9–
11). This higher temporal resolution is expected to decrease motion artifacts in patients with high heart rates as well as patients with arrhythmia (10,11).
Atrial fibrillation (AF) is the most common type of arrhythmia and the incidence increases markedly with ad- vancing age (12,13). AF is often associ- ated with structural heart disease. Be- fore initiating therapy, management of precipitating or reversible causes of AF is recommended (13). Coronary artery disease (CAD) is a cardiovascular con- dition associated with AF. Also, AF pa- tients may manifest symptoms mimick- ing CAD. Therefore, the demonstration of CAD is an important issue in the management of AF patients (13).
As motion artifacts, owing to high heart rates, particularly arrhythmia, impair image quality to a great extent, AF has remained as a contraindication for CT angiography applications. Multi- detector CT angiography was not con- sidered as a diagnostic tool in the clini- cal workflow of AF, but the improved temporal resolution achieved with dual- source CT may improve the visualiza- tion of coronary vessels in patients with AF. Thus, the purpose of our study was to evaluate the sensitivity and specificity of dual-source CT for significant coro- nary stenosis (⬎50% narrowing) in pa- tients with AF, by using conventional coronary angiography as the reference standard.
Materials and Methods
Patients
Institutional Review Board approval and informed consent was obtained from all patients.
Fifteen consecutive patients (nine men, six women; mean age⫾ standard deviation, 58.47 years ⫾ 9.07; range, 45–74 years) with AF were examined with dual-source CT between Septem- ber 2006 and January 2007. All patients were suspected of having coexisting CAD and were scheduled to undergo conventional angiography.
The exclusion criteria for CT were as follows: unstable clinical condition, previous allergy to iodinated contrast agents, elevated serum creatinine levels (⬎1.5 mg/dL, ⬎132.6 mol/L), previ- ous bypass surgery or stenting, and fail-
ure to follow breath-hold commands.
No patients were excluded on the basis of these criteria.
Conventional Coronary Angiography as Reference Standard
All patients underwent conventional cor- onary angiography with standard tech- niques by an experienced cardiologist (A.T., with 10 years coronary angiogra- phy experience) 1 day after CT exami- nation. The cardiologist who evaluated angiograms was blinded to the results of CT angiography. The coronary arteries were segmented according to the guide- lines of the American Heart Association (14). The vessel segments were evaluated by using the quantitative coronary analy- sis method (15). Lesions with a diame- ter reduction of 50% or more were con- sidered to be hemodynamically signifi- cant lesions. All coronary vessel segments were included.
CT Coronary Angiography: Scan Protocol and Reconstruction
All CT examinations were performed with a dual-source CT scanner (Soma- tom Definition; Siemens, Forchheim, Ger- many).
CT scans were obtained with colli- mation, 64 sections of 0.6-mm thick- ness; rotation time, 0.33 msec; tube voltage 120 kV; effective charge, 900
Published online 10.1148/radiol.2453070094 Radiology 2007; 245:703–711
Abbreviations:
AF⫽ atrial fibrillation CAD⫽ coronary artery disease CI⫽ confidence interval
LAD⫽ left anterior descending artery NPV⫽ negative predictive value PPV⫽ positive predictive value RCA⫽ right coronary artery Author contributions:
Guarantors of integrity of entire study, D.O., G.O.; study concepts/study design or data acquisition or data analy- sis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manu- script, all authors; literature research, D.O., G.O.; clinical studies, all authors; statistical analysis, D.O.; and manu- script editing, D.O., G.O.
Authors stated no financial relationship to disclose.
Advance in Knowledge
䡲 The overall sensitivity, specificity, positive predictive values, and negative predictive values of dual- source CT for significant coronary stenosis (⬎50% narrowing) in patients with atrial fibrillation (AF) were respectively calculated as 87%, 98%, 77%, and 99% for reader 1 and 80%, 99%, 80%, and 99% for reader 2.
Implications for Patient Care
䡲 Owing to improved temporal res- olution, dual-source CT shows potential for diagnostic noninva- sive coronary angiography in pa- tients with AF, previously a con- traindication for CT angiographic applications.
䡲 Dual-source CT may help pre- vent unnecessary invasive diag- nostic procedures to rule out coronary artery disease in pa- tients with AF.
mAs; pitch, 0.26 – 0.45 (depending on heart rate). The scan time was about 5.7– 8.1 seconds for a single breath hold and the scan was performed in a cranio- caudal direction. The patients did not re- ceive any additional drugs prior to the CT examination for heart rate regulation.
The CT angiography was triggered automatically by the arrival of the main contrast material– enhanced bolus (au- tomatic bolus tracking). We injected 70 mL nonionic contrast medium (Ultravist 350/mL; Iopromidum, Schering, Berlin, Germany) at a flow rate of 6 mL/sec.
This was followed by a 50-mL saline chaser bolus (5 mL/sec) to washout contrast from the right ventricle.
During the scan, electrocardiogra- phy was recorded simultaneously. The retrospective reconstructions were done in all cardiac phases with 50-msec inter- vals. For the reconstruction of transverse images, we used a section thickness of 0.75 mm and section width of 0.5 mm;
medium soft-tissue reconstruction ker- nels (B26f) were used. The reconstruc- tion interval with the fewest motion ar- tifacts was chosen and used for further analysis.
Prospective electrocardiographic tube current modulation (electrocardiographic pulsing) was not used to provide optimal radiation dose throughout the whole cardiac cycle.
Dual-source CT angiography was performed successfully without any com- plication in all 15 patients; all patients underwent conventional coronary an- giography.
CT Coronary Angiography Analysis The evaluation of the images was per- formed by two radiologists (D.O., G.O., with 5 years cardiac CT experience each), who were blinded to the results of conventional coronary angiography.
For the evaluation of coronary ar- tery stenosis, original transverse im- ages, multiplanar reformations, curved multiplanar reformations, maximum in- tensity projections, and volume-rendered images were used. For image postpro- cessing, a Leonardo (Siemens) worksta- tion was utilized.
The segmental evaluation of the cor- onary arteries was performed according
to the American Heart Association 15- segment classification. All segments were included for the assessment.
Image quality was evaluated qualita- tively for each coronary artery segment by using a three-point grading system:
grade 1, good image quality with no motion artifacts or minor artifacts not impairing the diagnostic quality; grade 2, moderate image quality with moder- ate artifacts and/or blurring but ade- quate and acceptable for clinical diagno- sis; and grade 3, poor or nonevaluable image quality with severe artifacts that made vessel delineation impossible. The segments with grade 3 were not consid- ered for further analysis. The reason for decreased image quality was considered for each segment.
The images were evaluated for the presence of significant stenosis. Simi- lar to conventional coronary angiogra- phy, significant stenosis was defined as a narrowing of the coronary lumen ex- ceeding 50%; all vessel segments were included in the analysis. Each vessel was analyzed on at least two planes, one parallel and one perpendicular to the course of the vessel. The vessel diameters were measured on recon- structions oriented perpendicular to the vessel course.
Comparisons and Evaluations
The results of CT coronary angiogra- phy to help detect significant stenoses (lesions ⬎50%) were compared with the results of conventional coronary angiography according to: (a) per-segment analysis, comparing each segment in every vessel; (b) per-vessel analysis, examining the presence of significant lesions in each of the major coronary vessels; and (c) per-patient analysis, evaluating the presence of any signifi- cant lesion in a given patient. Also, the diagnostic performance of dual-source CT coronary angiography to help de- tect significant stenosis in patients with AF was also evaluated with re- spect to image quality scores.
We also evaluated the patient dose. The dose-length product was displayed by the dual-source CT sys- tem itself and the dose-length product was then converted into effective dose
values by means of a conversion factor of 0.017 mSv/mGy䡠 cm, according to the Commission of the European Com- munities guidelines on quality criteria for CT (16).
Statistical Analysis
We did not perform power analysis to determine the number of patients in- cluded in the study because this repre- sented our initial experience. Statistical analysis was performed with statistical software (SPSS, version 12.0 for Win- dows; SPSS, Chicago, Ill). Sensitivity, specificity, PPV, and NPV were calcu- lated. These diagnostic parameters were expressed with a 95% confidence inter- val (CI). Conventional coronary angiog- raphy was regarded as the reference standard.
McNemar test was used to search for a significant difference between dual-source CT angiography and con- ventional coronary angiography to help detect coronary stenosis. A P value of less than .05 was considered to indicate a significant difference. To avoid bias caused by clustering effect resulting
Figure 1
Figure 1: Flow diagram shows study protocol.
from multiple observations in the same patients, the P value is corrected with an intracluster correlation factor (17).
Also, the results were checked with generalized estimating equation meth- ods (SAS/STAT, Release 8.2; SAS Insti- tute, Cary, NC).
Intermodality agreement between dual-source CT and conventional coro- nary angiography and interobserver agreement were determined by calcu- lating statistics. According to Landis and Koch (18), ⫽ 0 indicated poor agreement, ⫽ 0.01–0.20 indicated slight agreement, ⫽ 0.21–0.40 indi- cated fair agreement, ⫽ 0.41–0.60 indicated moderate agreement, ⫽ 0.61–
0.80 indicated good agreement, and ⫽ 0.81–1.00 indicated excellent agree- ment.
Results
Conventional Coronary Angiography A total of 225 vessel segments were an- alyzed (Fig 1). In fifteen (7%) of 225
segments with a diameter reduction of 50% or greater were found at conven- tional coronary angiography. No total occlusions were observed. The other 210 (93%) segments were found to be normal. No significant stenosis was found in six (40%) of 15 patients, whereas nine (60%) had CAD. Of those 15, conven- tional coronary angiography helped re- veal single-vessel disease in five (33%) patients, two-vessel disease in two (13%), and three-vessel disease in two (13%). Diseased segments with less than 50% narrowing were not consid- ered as positive results.
Dual-Source CT Angiography
The mean heart rate was 83.7 beats per minute ⫾ 8.9 (range, 69 –131 beats per minute) during CT angiogra- phy; all patients had irregular heart
Figure 2
Figure 2: CT angiography of 72-year-old woman with AF and no significant coronary artery stenosis. Heart rate was irregular (mean, 74 beats per minute; range 53–
142 beats per minute). Volume-rendered image reconstructed (a) at 300 msec after R wave shows poor image quality mimicking significant stenosis in RCA segment 2 (arrow), and (c) at 200 msec after R wave shows better image quality with no significant RCA stenosis observed. Sagittal maximum intensity projection image of RCA reconstructed (b) at 300 msec after R wave shows poor image quality with false impression of stenotic lesion in segment 2 (arrow), and (d) at 200 msec after R wave shows better image image quality with no stenosis. (e) Left anterior oblique angiographic view confirms normal RCA.
Table 1
Image Quality Grading of Dual-Source CT Coronary Angiography
Image Quality Reader 1 Reader 2 Good (grade 1) 101 (45) 103 (46) Moderate (grade 2) 108 (48) 109 (48) Poor (grade 3) 16 (7) 13 (6)
Total 225 225
Note.—Numbers are coronary segments graded. Num- bers in parentheses are percentages.
rates. In all of the CT angiography im- ages, the reconstructions were done cov- ering the whole cardiac cycle with abso- lute intervals in milliseconds. In all patients, the reconstructions obtained during systole had fewer motion artifacts, and thus better image quality. In 10 of the study group patients, 300 msec (after R wave) was the best interval with the few- est artifacts. The 250-msec intervals in three patients and 200-msec intervals in two patients demonstrated the fewest artifacts and were used for further anal- ysis (Fig 2). Both readers used the same data sets for further evaluation.
The mean scan time was 6.64 sec- onds⫾ 0.74 (range, 5.7–8.1 seconds).
The mean scan pitch was 0.35 second⫾ 0.06 (range 0.26 – 0.45 second). The av- erage patient dose was 13.8 mSv⫾ 1.37 (range, 11.50 –15.9 mSv).
In 15 patients, a total of 225 seg- ments were evaluated for image qual- ity with respect to vessel visibility and existing artifacts. Reader 1 (D.O.) evaluated 101 (45%) segments as good, 108 (48%) as moderate, and 16 (7%) as poor image quality; whereas reader 2 (G.O.) evaluated 103 (46%) segments as good, 109 (48%) as mod- erate, and 13 (6%) as poor image quality (Table 1). The segments with poor image quality were not consid- ered for further analysis.
For reader 1, the reason for de- creased image quality was residual car- diac motion in 13 patients and heavy wall calcification in three. For reader 2, the reason for decreased image quality was residual cardiac motion in 11 pa-
tients and heavy wall calcification in two.
Motion artifacts resulted in blurred or doubled vessel contours. Most mo- tion artifacts affected the middle section of the RCA and the left circumflex coro-
nary artery. The left main coronary ar- tery and proximal left anterior descend- ing artery were the least affected seg- ments with the best image quality.
Regarding the false-positive and false- negative results, in all of the misdiag- Figure 3
Figure 3: CT angiography of 67-year-old man with AF and three-vessel disease. Heart rate was irregular (mean, 82 beats per minute; range, 67–110 beats per minute). (a) Curved multiplanar reformation image of RCA shows two stenoses in segment 3 (arrows). Note multiple atherosclerotic plaques leading to milder luminal narrowing (arrowheads). Image quality was good; images reconstructed at 300 msec after R wave.
(b)Sagittal maximum intensity projection shows stenosis in RCA segment 3 (arrows) with additional athero- sclerotic changes (arrowheads). (c) Volume-rendered image shows stenoses and accompanying atheroscle- rotic lesions in RCA segment 3. (d) Left anterior oblique angiogram shows lesions in distal RCA segment.
Table 2
Effectiveness of Dual-Section CT to Demonstrate Coronary Artery Stenosis in Patients with AF
Analysis and Reader No. Sensitivity (%) Specificity (%) PPV (%) NPV (%)
Per segment
1 (n⫽ 209) 8 (13/15) [70, 100] 98 (190/194) [96, 99] 77 (13/17) [59, 96] 99 (190/192) [98, 100]
2 (n⫽ 212) 80 (12/15) [60, 100] 99 (194/197) [97, 100] 80 (12/15) [60, 100] 99 (194/197) [97, 100]
Per vessel
1 (n⫽ 60) 87 (13/15) [70, 100] 96 (43/45) [90, 100] 87 (13/15) [70, 100] 96 (43/46) [90, 100]
2 (n⫽ 60) 80 (12/15) [60, 100] 96 (43/45) [89, 100] 93 (12/14) [86, 100] 92 (43/46) [84, 99]
Per patient
1 (n⫽ 15) 89 (8/9) [68, 100] 83 (5/6) [54, 100] 89 (8/9) [68, 100] 83 (5/6) [54, 100]
2 (n⫽ 15) 100 (7/7) 75 (6/8) [50, 100] 78 (7/9) [50, 100] 100 (6/6)
Note.—Numbers in parentheses are evaluated segments. Numbers in square brackets are 95% CIs.
noses, the image quality was moderate.
No segment with good image quality was misdiagnosed by both readers. Sen-
sitivity, specificity, PPV, and NPV of dual-source CT angiography to help de- tect stenoses in patients with AF were
all 100% when the image quality was good ( Figs 3, 4).
When the images were degraded
Table 3
Effectiveness of Dual-Source CT Coronary Angiography to Help Detect Coronary Stenosis in AF Patients Coronary Artery Segment and Reader No.*
Result Statistic†
True-Positive True-Negative False-Positive False-Negative Sensitivity (%) Specificity (%) PPV (%) NPV (%) Right main
1 (n⫽ 54) 5 47 1 1 71 (38, 100) 98 (94, 100) 83 (54, 100) 96 (90, 100)
2 (n⫽ 55) 4 46 3 2 67 (29, 100) 94 (87, 100) 57 (20, 94) 96 (90, 100)
Proximal [AHA 1]
1 (n⫽ 15) 2 13 0 0 100 100 100 100
2 (n⫽ 15) 2 13 0 0 100 100 100 100
Middle [AHA 2]
1 (n⫽ 12) 1 10 1 0 100 91 (74, 100) 50 (0, 100) 100
2 (n⫽ 13) 0 9 2 1 0 82 (59, 100) 0 90 (71, 100)
Distal [AHA 3]
1 (n⫽ 12) 2 9 1 1 67 (13, 100) 100 100 90 (70, 100)
2 (n⫽ 12) 2 8 0 1 67 (13, 100) 89 (68, 100) 67 (13, 100) 89 (68, 100)
Posterior descending [AHA 4]
1 (n⫽ 15) 0 15 0 0 NA 100 NA 100
2 (n⫽ 15) 0 15 0 0 NA 100 NA 100
Left main [AHA 5]
1 (n⫽ 15) 0 15 0 0 NA 100 NA 100
2 (n⫽ 15) 0 15 0 0 NA 100 NA 100
Left anterior descending
1 (n⫽ 71) 6 62 2 1 86 (60, 100) 97 (93, 100) 75 (45, 100) 98 (95, 100)
2 (n⫽ 72) 7 65 0 0 100* 100 100 100
Proximal [AHA 6]
1 (n⫽ 15) 1 14 0 0 100 100 100 100
2 (n⫽ 15) 1 14 0 0 100 100 100 100
Middle [AHA 7]
1 (n⫽ 14) 3 10 0 1 75 (33, 100) 100 100 91 (74, 100)
2 (n⫽ 15) 4 11 0 0 100 100 100 100
Distal [AHA 8]
1 (n⫽ 14) 0 13 1 0 NA 93 (79, 100) NA 100
2 (n⫽ 14) 0 14 0 0 NA 100 NA 100
First diagonal branch [AHA 9]
1 (n⫽ 15) 1 13 1 0 100 93 (79, 100) 100 93 (80, 100)
2 (n⫽ 15) 1 14 0 0 100 100 100 100
Second diagonal branch [AHA 10]
1 (n⫽ 13) 1 12 0 0 100 100 100 100
2 (n⫽ 13) 1 12 0 0 100 100 100 100
Left circumflex artery
1 (n⫽ 69) 2 66 1 0 100 99 (96, 100) 67 (13, 100) 100
2 (n⫽ 70) 1 68 0 1 50 (20, 100) 100 100 99 (96, 100)
Proximal [AHA 11]
1 (n⫽ 15) 1 14 0 0 100 100 100 100
2 (n⫽ 15) 1 14 0 0 100 100 100 100
Middle [AHA 13]
1 (n⫽ 13) 1 12 0 0 100 100 100 100
2 (n⫽ 14) 0 13 0 1 0 100 NA 93 (79, 100)
Distal [AHA 15]
1 (n⫽ 13) 0 13 0 0 NA 100 NA 100
2 (n⫽ 13) 0 13 0 0 NA 100 NA 100
(Table 3 continues)
(moderate image quality), sensitivity, specificity, PPV, and NPV were re- spectively calculated as 71% (five of seven, 95% CI: 31%, 100%), 96% (97 of 101, 95% CI: 94%, 98%), 56% (five of nine, 95% CI: 20%, 93%), and 98%
(97 of 99, 95% CI: 96%, 100%) for reader 1 and 63% (five of eight, 95%
CI: 25%, 98%), 97% (98 of 101, 95%
CI: 95%, 99%), 63% (five of eight, 95% CI: 25%, 98%), and 97% (98 of 101, 95% CI: 95%, 99%) for reader 2.
The respective overall sensitivity, specificity, PPV, and NPV were 87%
(13 of 15, 95% CI: 70%, 100%), 98%
(190 of 194, 95% CI: 96%, 99%), 77% (13 of 17, 95% CI: 56%, 96%), and 99% (190 of 192, 95% CI: 98%, 100%) for reader 1 and 80% (12 of 15, 95% CI: 60%, 100%), 99% (194 of 197, 95% CI: 97%, 100%), 80% (12 of 15, 95% CI: 60%, 100%), and 99%
(194 of 197, 95% CI: 97%, 100%) for reader 2.
Per-segment analysis.—Reader 1 cor- rectly identified 13 of 15 significant lesions with a diameter reduction of 50% or greater. Two stenoses were missed (RCA segment 3 and LAD seg- ment 7). Four lesions were detected at dual-source CT only and were not con- firmed with conventional coronary an- giography and were thus considered as false-positives (RCA segment 2, LAD segments 8 and 9, and left cir- cumflex artery segment 13).
Reader 2 correctly diagnosed 12 of
Figure 4
Figure 4: CT angiography of 44-year-old man with AF and single-vessel disease. Heart rate was irreg- ular (mean, 77 beats per minute; range, 69 –98 beats per minute). (a) Maximum intensity projection image shows stenosis in LAD segment 6. Image quality was good; images reconstructed at 300 msec after R wave. (b) Curved multiplanar reformation image shows stenosis in LAD segment 6 (arrow). Note atherosclerotic plaques leading to contour irregularities (arrowheads). Discontinuity readings in distal LAD and diagonal branch segments resulting from arrhythmia do not complicate interpretation. (c) Vol- ume-rendered image shows stenosis in LAD segment 6, discontinuity in distal segments. (d) Right ante- rior oblique angiographic view shows stenosis in LAD segment 6 (arrow), atherosclerotic contour irreg- ularities (arrowheads).
Table 3 (continued)
Effectiveness of Dual-Source CT Coronary Angiography to Help Detect Coronary Stenosis in AF Patients
Coronary Artery Segment and Reader No.*
Result Statistic†
True-Positive True-Negative False-Positive False-Negative Sensitivity (%) Specificity (%) PPV (%) NPV (%) First obtuse media branch [AHA 12]
1 (n⫽ 14) 0 13 1 0 NA 93 (79, 100) NA 100
2 (n⫽ 14) 0 14 0 0 NA 100 NA 100
Second obtuse media branch [AHA 14]
1 (n⫽ 14) 0 14 0 0 NA 100 NA 100
2 (n⫽ 14) 0 14 0 0 NA 100 NA 100
Total
1 (n⫽ 209) 13 190 4 2 87 (70, 100) 98 (96, 99) 77 (56, 96) 99 (98, 100)
2 (n⫽ 212) 12 194 3 3 80 (60, 100) 99 (97, 100) 80 (60, 100) 99 (97, 100)
* American Heart Association segment number.
†Numbers in parentheses are 95% CIs.
NA⫽ not applicable.
15 important lesions. Three stenoses were misdiagnosed as normal (RCA segments 2 and 3 and left circumflex artery segment 12) and three normal segments were misdiagnosed as steno- ses (RCA segments 2 and 3 and RCA segment 3).
Per-vessel analysis.—Reader 1 misdi- agnosed two vessels without CAD as diseased (false-positive) and two vessels with stenoses as normal (false-nega- tive).
Reader 2 misdiagnosed two vessels without CAD as diseased (false-posi- tive) and three vessels with stenoses as normal (false-negative).
Per-patient analysis.—Reader 1 misdi- agnosed single-vessel CAD in one pa- tient as normal (false-negative) and ab- sence of CAD as single-vessel CAD (false-positive) in one patient.
Reader 2 misdiagnosed single-vessel CAD (false-positive) in two normal pa- tients. CAD was not misdiagnosed as normal in any patient.
The sensitivity, specificity, PPV, and NPV values of dual-source CT angiogra- phy for segment-, vessel-, and patient- based analysis for both readers are given in Table 2, while Table 3 demon- strated the diagnostic performance for each coronary artery and segment sepa- rately.
With McNemar test, no significant difference between dual-source CT an- giography and conventional coronary angiography was found in helping detect significant stenosis (reader 1, corrected P ⫽ .31 and reader 2, corrected P ⫽ .29). The analysis with generalized esti- mating equation methods verified our results.
Additionally, statistics demonstrated good intermodality agreement for both readers (reader 1, ⫽ 0.79 and reader 2, ⫽ 0.78).
The interobserver agreement in de- tecting significant stenosis was calcu- lated as ⫽ 0.865, corresponding to excellent agreement.
Discussion
In our study, we obtained the images with highest quality in the reconstruc- tions performed at end-systolic phases.
Usually, in patients with sinus rhythm and low heart rates, good image quality is achievable with reconstructions per- formed at around 70%–75% of the R-R interval, corresponding to the mid dia- stolic slow-filling phase in which cardiac motion is slowest (19–22). However, in patients with AF, although the R-R in- terval varies in each cardiac cycle, the variation between end-diastolic time to end-systolic time and that of the fast- inflow phase in early diastole is small, whereas the variation in the subsequent slow-inflow phase (mid diastole) is large (19–22). Therefore, during mid dias- tole, the temporal variation differs sig- nificantly between cardiac cycles.
On the other hand, at end-systole, the temporal variation in each cardiac cycle is small; therefore, the image qual- ity is better. This results in superior image quality of the images recon- structed at end systole rather than those reconstructed at mid diastole in patients with AF. This is also why absolute inter- vals are preferred for reconstructions instead of relative intervals.
Despite the high temporal resolu- tion and appropriate selection of the re- construction intervals, the main source of image-degrading artifacts was resid- ual cardiac motion. The RCA appears particularly vulnerable because of its ex- tensive motion radius and short motion- free period (21,22). In our study, the middle sections of the RCA and the left circumflex artery were the most af- fected segments. The left main coronary artery and proximal LAD were the least affected segments with highest image quality, possibly owing to their horizon- tal course.
Apart from blurring or double-con- tour appearance owing to residual heart motion, some vessel discontinuity also resulted. However, this discontinuity was not overly severe in most cases and did not complicate image interpretation to a great extent. No breath-holding prob- lems occurred, owing to short scan times.
Extensive vessel wall calcifications also impaired image quality and made interpretation difficult, especially when associated with residual cardiac motion artifacts. In the case of calcifications, we
used multiplanar reformation and curved multiplanar reformation images for evaluation. Although the presence of calcifications made the evaluation diffi- cult, they did not lead to false results.
Considering the false-positive and false-negative results, no segment with good image quality was misdiagnosed by both readers. Therefore, when image quality was good, the performance of dual-source CT to help detect coronary stenosis, even in patients with AF, was excellent.
Prospective electrocardiographically gated tube current modulation (electro- cardiographic pulsing), which is ap- plied to minimize the patient dose by limiting the optimal radiation dose to a certain period of the cardiac cycle (usually to diastolic phase) (23,24), was not used in our study to allow maximal flexibility in reconstruction intervals. With faster heart rates, the optimal timepoint for image recon- struction becomes more difficult to predict and frequently occurs during the end-systolic phase of total myocar- dial contraction (25–28). In AF pa- tients the R-R intervals demonstrate great between-beat variability. This increased the patient dose accord- ingly, but heart rate–adaptive table feed settings helped decrease the pa- tient dose to some extent by shorten- ing acquisition time. This property also helped obtain more motion-free images at high heart rates (10,11,29).
Although promising results were ob- tained, the main limitation of our study was the small number of patients in- cluded in our study group. However, this represented our initial experience and this low number of patients is not enough to draw conclusions. For exam- ple, only one false-negative evaluation resulted in a low NPV for reader 1 in per-patient analysis. The small number of patients resulted in a small number of true-negative evaluations (n⫽ 5, NPV ⫽ five of six).
Also, the results were influenced by the high prevalence of CAD in our patient group because they were all suspected of having CAD and were scheduled to un- dergo conventional coronary angiogra- phy. Thus, the sensitivity and specificity
values may differ in AF patients with a lower prevalence of CAD. Our study also did not include semiquantitative analysis of coronary CT angiography to provide more objective criteria.
In conclusion, the recently introduced dual-source CT technology provides bet- ter temporal resolution and minimizes motion artifacts, which allows coronary angiography at higher heart rates, even in cases of arrhythmia. The high NPV may be useful for obviating invasive coronary angiography in patients with AF whose symptoms or abnormal stress test results make it necessary to rule out the pres- ence of coronary artery stenosis to guide therapy. Therefore, dual-source CT has the potential to make noninvasive coro- nary angiography effective in a signifi- cantly increased number of patients and in a wider spectrum of clinical situations as compared with earlier scanners.
Larger studies will be needed to deter- mine if our results are reproducible.
References
1. Schoepf UJ, Becker C, Ohnesorge BM, Yucel K.
CT of coronary artery disease. Radiology 2004;
232:18–37.
2. Schoenhagen P, Halliburton SS, Stillman AE, et al. Noninvasive imaging of coronary arteries: current and future role of multi–
detector row CT. Radiology 2004;232:7–17.
3. Achenbach S, Giesler T, Ropers D, et al.
Detection of coronary artery stenoses by contrast-enhanced, retrospectively electro- cardiographically-gated, multislice spiral computed tomography. Circulation 2001;
103:2535–2538.
4. Nieman K, Cademartiri F, Lemos PA, Raaijmakers R, Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiog- raphy with fast submillimeter multislice spiral computed tomography. Circulation 2002;106:2051–2054.
5. Heuschmid M, Kuettner A, Schroeder S, et al. ECG-gated 16-MDCT of the coronary arteries: assessment of image quality and ac- curacy in detecting stenosis. AJR Am J Roentgenol 2005;184:1413–1419.
6. Pugliese F, Mollet NR, Runza G, et al. Diag- nostic accuracy of non-invasive 64-slice CT coronary angiography in patients with stable angina pectoris. Eur Radiol 2006;16(3):575–
582.
7. Raff GL, Gallagher MJ, O’Neill WW, Goldstein JA. Diagnostic accuracy of non- invasive coronary angiography using 64- slice spiral computed tomography. J Am Coll Cardiol 2005;46(3):552–557.
8. Sun Z, Jiang W. Diagnostic value of multi- slice computed tomography angiography in coronary artery disease: a meta-analysis.
Eur J Radiol 2006;60(2):279 –286.
9. Flohr TG, McCollough CH, Bruder H, et al.
First performance evaluation of a dual- source CT (DSCT) system. Eur Radiol 2006;
16(2):256 –268.
10. Achenbach S, Ropers D, Kuettner A, et al. Con- trast-enhanced coronary artery visualization by dual-source computed tomography: initial expe- rience. Eur J Radiol 2006;57(3):331–335.
11. Johnson TR, Nikolaou K, Wintersperger BJ, et al. Dual-source CT cardiac imaging: initial experience. Eur Radiol 2006;16(7):1409 – 1415.
12. Chatap G, Giraud K, Vincent JP. Atrial fibril- lation in the elderly: facts and management.
Drugs Aging 2002;19(11):819 – 846.
13. Fuster V, Ryden LE, Cannom DS, et al. ACC, AHA, ESC 2006 Guidelines for the Manage- ment of Patients with Atrial Fibrillation: a report of the American College of Cardiology and American Heart Association Task Force on Practice Guidelines and the European So- ciety of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients With Atrial Fibrillation)— devel- oped in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society. Circulation 2006;114(7):
e257– e354.
14. Austen WG, Edwards JE, Frye RL, et al. A reporting system on patients evaluated for coronary artery disease: report of the Ad Hoc Committee for Grading of Coronary Ar- tery Disease, Council on Cardiovascular Sur- gery, American Heart Association. Circula- tion 1975;51(suppl 4):5– 40.
15. Leber AW, Knez A, Ziegler FZ, et al. Quantifica- tion of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrasound. J Am Coll Cardiol 2005;46(1):147–152.
16. Commission of the European Communities.
European guidelines on quality criteria for computed tomography: EUR 16262EN.
Luxembourg: Office for the Official Publica- tions of the European Communities, 2000.
17. Gonen M, Panageas KS, Larson SM. Statisti- cal Issues in analysis of diagnostic imaging
experiments with multiple observations per patient. Radiology 2001;221:763–767.
18. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Bi- ometrics 1977;33:159 –174.
19. Hoffmann MH, Shi H, Manzke R, et al. Non- invasive coronary angiography with 16-de- tector row CT: effect of heart rate. Radiology 2005;234:86 –97.
20. Vembar M, Garcia MJ, Heuscher DJ, et al. A dynamic approach to identifying desired physiological phases for cardiac imaging us- ing multislice spiral CT. Med Phys 2003;30:
1683–1693.
21. Wang Y, Vidan E, Bergman GW. Cardiac motion of coronary arteries: variability in the rest period and implications for coronary MR angiography. Radiology 1999;213:751–
758.
22. Achenbach S, Ropers D, Holle J, Muschiol G, Daniel WG, Moshage W. In-plane coro- nary arterial motion velocity: measurement with electron-beam CT. Radiology 2000;216:
457– 463.
23. Jakobs TF, Becker CR, Ohnesorge B, et al.
Multislice helical CT of the heart with retro- spective ECG gating: reduction of radiation exposure by ECG-controlled tube current modulation. Eur Radiol 2002;12(5):1081–
1086.
24. Abada HT, Larchez C, Daoud B, Sigal- Cinqualbre A, Paul JF. MDCT of the coro- nary arteries: feasibility of low-dose CT with ECG-pulsed tube current modulation to re- duce radiation dose. AJR Am J Roentgenol 2006;186(6 suppl 2):S387–S390.
25. Stein PD, Beemath A, Kayali F, Skaf E, Sanchez J, Olson RE. Multidetector com- puted tomography for the diagnosis of coro- nary artery disease: a systematic review.
Am J Med 2006;119:203–216.
26. Herzog C, Arning-Erb M, Zangos S, et al. Multi- detector row CT coronary angiography: influ- ence of reconstruction technique and heart rate on image quality. Radiology 2006;238:75–86.
27. Leschka S, Wildermuth S, Boehm T, et al.
Noninvasive coronary angiography with 64- section CT: effect of average heart rate and heart rate variability on image quality. Radi- ology 2006;241:378 –385.
28. Kovacs A, Probst C, Sommer T, et al. CT coronary angiography in patients with atrial fibrillation [in German]. Rofo 2005;177(12):
1655–1662.
29. Scheffel H, Alkadhi H, Plass A, et al. Accuracy of dual-source CT coronary angiography: first ex- perience in a high pre-test probability population without heart rate control. Eur Radiol 2006;
16(12):2739–2747.
