- Email: [email protected]
Assessment of Coronary Artery Stent Patency and Restenosis Using 64-Slice Computed Tomography
Assessment of Coronary Artery Stent Patency and Restenosis Using 64-Slice Computed Tomography1 Carsten Rist, Franz von Ziegler, Konstantin Nikolaou, Miles A. Kirchin, Bernd J. Wintersperger, Thorsten R. Johnson Andreas Knez, Alexander W. Leber, Maximilian F. Reiser, Christoph R. Becker
Rationale and Objectives. Restenosis remains a major limitation of coronary catheter– based stent placement. Therefore, a reliable noninvasive diagnostic method for the evaluation of stented coronary arteries would be highly desirable. Our aim was to evaluate the diagnostic accuracy of high-resolution 64-slice computed tomography (64SCT) in a pilot study for the assessment of the lumen of coronary artery stents. Materials and Methods. Twenty-five patients underwent 64SCT of the coronary arteries and quantitative x-ray coronary angiography (QCA) after coronary artery stent placement. 64SCT coronary angiography was performed with the following parameters: spatial resolution ⫽ 0.4 ⫻ 0.4 ⫻ 0.4 mm; temporal resolution ⫽ 83–165 milliseconds; contrast agent ⫽ 80 mL at a flow rate of 5 mL/second; retrospective electrocardiogram gating. The 64SCT scans were evaluated for image quality and for the presence of significant in-stent and peri-stent (proximal and distal) stenoses. Determinations were made of the sensitivity, specificity, diagnostic accuracy, and positive and negative predictive values (PPV and NPV) of 64SCT for the detection or exclusion of stenoses. Results. A total of 46 stents were evaluated, of which 45 (98%) were of diagnostic image quality. Significant in-stent restenosis or occlusion was detected on QCA in 8/45 cases (ⱖ50% stenosis ⫽ 6; occlusion ⫽ 2). The sensitivity, specificity, accuracy, PPV, and NPV of 64SCT for the detection of significant in-stent disease was 75%, 92%, 89%, 67%, and 94%, respectively. Both occluded coronary artery stents were correctly identified. The sensitivity, specificity, and accuracy values of 64SCT for the detection of significant proximal peri-stent stenoses were 75%, 95%, and 93%, respectively, whereas the values for detection of significant distal peri-stent stenoses were 67%, 85%, and 84%, respectively. Conclusion. The high spatial and temporal resolution of 64SCT may permit improved assessment of stent occlusion and peri-stent disease, although detection of in-stent stenosis remains difficult. Key Words. Coronary artery disease; multislice computed tomography; atherosclerosis; coronary artery stent. ©
AUR, 2006
Coronary artery stenting is the predominant method of nonsurgical myocardial revascularization; an estimated 475,000 coronary artery stents were implanted in the United States alone in 2001 (1). Unfortunately, de novo coronary lesions
Acad Radiol 2006; 13:1465–1473 1
From the Department of Clinical Radiology (C.R., K.N., B.J.W., T.J., M.F.R., C.R.B.) and Department of Cardiology (F.v.Z., A.K., A.W.L.), University Hospitals – Grosshadern Ludwig-Maximilians University, Marchioninistr. 15, 81377 Munich, Germany, and Worldwide Medical Affairs, Bracco Imaging SpA, Milan, Italy (M.A.K.). Received June 29, 2006; accepted September 8, 2006. C.R. and F.v.Z. have contributed equally to this article. Address correspondence to: C.R. e-mail: [email protected]
© AUR, 2006 doi:10.1016/j.acra.2006.09.044
and restenosis are major problems in a considerable proportion of patients; a 6-month restenosis rate ranging from 11% to 46% has been reported if non– drug-eluting stents are used (2,3). Patients with restenosis often require repeat interventional procedures that result in a reduced overall quality of life and increased costs to the health care system (4). Initial clinical experience with drug-eluting coronary stents has shown excellent results, with 0% restenosis at 4-month, 6-month, and 12-month follow-up examinations (5,6). Nevertheless, despite these recent findings, restenosis remains the major limitation of coronary catheter-based intervention. A reliable noninvasive diagnostic method for the evaluation of stented coronary arteries would therefore be highly desirable.
1465
RIST ET AL
The introduction of multislice computed tomography (MSCT) technology combined with fast rotation time and retrospective electrocardiogram gating permits the entire heart to be examined in a single breath hold, while maintaining high submillimeter isotropic spatial resolution (7). However, whereas the important role of 16-slice CT systems (16SCT) in MSCT coronary angiography for the diagnosis of coronary artery disease is well-established (8), the potential benefits for evaluation of coronary artery stents has still to be established. Early studies using a four-slice CT scanner revealed frequent false-positive diagnoses of high-grade in-stent stenoses because of artificial thickening of the stent struts (9). More recent in vitro experiments on a 16SCT system to assess different stent types in terms of stent lumen visualization revealed that visualization was largely dependent on the metallic composition and diameter of the particular stent (10). Promising results for the visualization of stent lumen have subsequently been obtained with 16SCT and with 40/64-slice (64SCT) for the detection of significant in-stent stenosis (11–13). The purpose of our study was to investigate the potential of MSCT coronary angiography using a 64SCT system for the assessment of stented coronary arteries. Specifically, our study was aimed at evaluating the sensitivity, specificity, and diagnostic accuracy of 64SCT coronary angiography for the detection of significant instent and peri-stent steno-occlusive disease.
MATERIAL AND METHODS Patient Population The study population comprised 25 consecutive patients (23 male, 2 female; mean age 59.4 ⫾ 12 years; range 40 – 83 years). Patients were included over a period of 4 months. Eligible patients had each undergone percutaneous transluminal coronary angioplasty with stent placement. After this interventional procedure, all patients underwent a second conventional coronary angiography and a comparative MSCT angiography of the coronary arteries. The mean time interval between MSCT and this second conventional angiography was 5 ⫾ 2 days (range 1–9 days), whereas MSCT was performed at a mean interval of 22 ⫾ 41.8 days (range 1–182 days) after coronary stent placement. For the majority of our study population (n ⫽ 17), the indication for conventional angiography after stent placement was the suspicion of in-stent restenosis because of new or changing clinical symptoms.
1466
Academic Radiology, Vol 13, No 12, December 2006
For the remaining patients (n ⫽ 8), the indication for catheter-based coronary angiography was an Institutional Review Board–approved clinical study to evaluate drugeluting and drug-coated stents versus non– drug-eluting stents over a 3-month period. Exclusion criteria were 1) atrial fibrillation and other severe arrhythmias, 2) severe renal failure, 3) hyperthyroidism, 4) hypersensitivity to iodine-containing contrast agent, 5) pregnancy, and 6) unstable clinical conditions. The mean heart rate during the MSCT examinations was 62.2 ⫾ 8 bpm (range 49 –79 bpm). Beta-blocker treatment (Metoprolol 5–10 mg) was administered intravenously to patients (n ⫽ 14) with heart rates ⬎70 bpm shortly before the MSCT examination. Metoprolol was not administered to patients with known hypersensitivity or contraindications for beta-blockers. All patients gave informed consent to participate in the study. The study protocol was approved by the Ethics Committee of the hospital. Stent Characteristics A total of 46 stents were implanted in the 25 patients included in the study. Nine different stent types were employed; the material and design of these stents are summarized in Table 1. The majority of the stents used (27/46) were made of surgical stainless steel (316 L). The remaining 19 stents comprised 6 made of chrome-cobalt composition, 7 that were paclitaxel-eluting/coated stents, and 6 that were rapamycin-eluting/coated stents. The diameter of the implanted stents ranged from 2.5 mm to 4.0 mm. The site of stent implantation was the right coronary artery in 17 cases, the left main in 1 case, the left anterior descending in 20 cases, and the left circumflex in 8 cases (Table 2). MSCT Scanning Protocol MSCT was performed using a new-generation 64SCT system (Somatom Sensation 64, Siemens Medical Solutions, Forchheim, Germany) working at an increased rotation rate and with application of z-axis flying focal spot technology to achieve an effective slice-thickness of 0.6 mm (14). A contrast medium bolus of 80 mL (Iomeron 300, Bracco Imaging SpA, Milan, Italy) was injected into the antecubital vein at a flow rate of 5 mL/second. Automated bolus tracking software was employed to time the start of the scan to the arrival of the contrast medium bolus; each scan began automatically 6 seconds after the contrast density in the ascending aorta reached a predefined threshold of 100 Hounsfield units (15). Images
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Table 1 Properties of Balloon-Expandable Stents (n ⴝ 46) and Commercial Name of Stents, Manufacturer, Material, Length, and Nominal Diameter of the Examined Stents
Name Paclitaxel MAC Arthos Bx-Sonic Cypher select Arthos Pico Flex Master Bx-Velocity Taxus
Manufacturer
Material
Cook Mönchengladbach, Germany AMG Raesfeld-Erle, Germany AMG Raesfeld-Erle, Germany Cordis Langenfeld, Germany Cordis Langenfeld, Germany
Stainless steel 316 L Paclitaxel Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Rapamycin Chrome-Cobalt composition Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Paclitaxel
AMG Raesfeld-Erle, Germany Jomed, Haan, Germany Cordis Langenfeld, Germany Boston Scientific Ratingen, Germany
Table 2 Distribution of Implanted Coronary Artery Stents According to the AHA-15 Segment Coronary Artery Model
RCA
LM LAD
Rd1 Rd2 LCx
Total
Segment
Number of Stents
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
5 7 5 0 1 7 11 0 2 0 6 1 1 0 0 46
AHA, American Heart Association; RCA, right coronary artery; LM, left main artery; LAD, left anterior descending artery; Rd1, first diagonal branch; Rd2, second diagonal branch; LCx, left circumflex artery.
were acquired during an inspiratory breath-hold; the entire volume of the heart was covered in about 10 seconds. The electrocardiogram signal of the patient was recorded simultaneously during the examination to enable retrospective gating of the data. The gantry rotation time was 330 milliseconds, the tube current was 850 mAs using dose modulation (16), and the tube voltage was 120 kV.
Length (mm)
Strut Dimensions (mm)
Diameter (mm)
Number of Implanted Stents
16
0.125
2.5, 3.5
3
9, 13 33 18 33
0.125 0.125 0.14 0.14
3.5, 4.0 3 3 3.5
6 1 2 6
14, 19 9 8, 13 16
0.065–0.075 0.09 0.14 n/a
2.5, 3.0 3.0 3.0 3.0
6 13 5 4
The temporal resolution was 83–165 milliseconds, the detector collimation was 64 ⫻ 0.6 mm, and the spatial resolution was 0.4 ⫻ 0.4 ⫻ 0.4 mm. For optimal, motion-free image quality, datasets were reconstructed in mid-diastole (mean interval 614 ⫾ 175 milliseconds after the R-wave). Based on these protocol parameters the estimated radiation dose was approximately 8 –10 mSv. Most datasets were reconstructed using a medium smooth body convolution kernel (B30f) as well as a sharp heart view (B46f) convolution kernel. The B46f kernel was used in datasets for 19/25 patients (34/45 stents). Conventional X-Ray Coronary Angiography Both conventional x-ray coronary angiography and balloon angioplasty with stenting procedures were performed according to standard techniques. Coronary angiograms were evaluated quantitatively (QuantCor QCA, Siemens Medical Systems) by an independent investigator who was fully blinded to the findings of MSCT angiography. After visual assessment, quantitative x-ray coronary angiography (QCA) was performed of all stented vessels and proximal and distal coronary artery lumina. The mean diameter reduction of these lesions in two projections was determined and the segmental location of the stenoses according to the 15-segment AHA (American Heart Association model) was recorded (17). MSCT Image Analysis All MSCT datasets were transferred to a commercially available postprocessing workstation (Leonardo, Siemens)
1467
RIST ET AL
and analyzed by two independent experienced readers who were blinded to the results of x-ray coronary angiography. Images were evaluated initially for diagnostic quality by both readers in consensus. A 3-point scale was used: 1 ⫽ high image quality (optimal depiction of the coronary artery and the stented area); 2 ⫽ moderate image quality (sufficient depiction of coronary artery and coronary artery stent for diagnostic purposes, but minor motion artifacts or reduced signal-to-noise ratio); and 3 ⫽ poor or uninterpretable image quality (data not adequate for diagnostic assessment of stent lumina or proximal and distal stent areas). Images considered to be of high or moderate diagnostic quality were then evaluated further by the two readers independently. Evaluation was performed of the original axial source images and of postprocessed curved multiplanar reformations, maximum intensity projection reconstructions, and volume rendered three-dimensional reconstructions. Each coronary artery stent segment was independently evaluated according to a four-point scale, in which 1 ⫽ patent stent lumen, no in-stent disease; 2 ⫽ stent lumen patent, but in-stent restenosis of ⬍50% of the vessel lumen diameter due to atherosclerotic vessel wall changes or intimal hyperplasia; 3 ⫽ significant in-stent restenosis of ⱖ50% of the vessel lumen diameter; and 4 ⫽ stent occlusion. A stent lumen was considered to be patent if contrast medium was detected inside the stent lumen and if peripheral enhancement was present. Finally, independent evaluation was also performed to determine the presence/absence of stenosis 3 mm proximal and distal to the stent (peri-stent disease). Narrowing of the luminal diameter by ⱖ50% was considered to indicate significant proximal or distal stent stenosis. After the independent reading, a consensus reading was performed to obtain a final MSCT diagnosis. Statistical Analysis Statistical analysis was performed using two Windowsbased software products [MedCalc, Version 7.0.0.2, 2002 (Mariakerke, Belgium); SPSS 12.0.1, 2003 (Chicago, Illinois)]. Continuous variables were presented as mean ⫾ standard deviation. The value of 64SCT for the detection of significant coronary artery stent restenosis was determined against conventional angiography as the standard of reference. The sensitivity, specificity, diagnostic accuracy, and positive and negative predictive values (PPV and NPV, respectively) were calculated. A significance level of P ⬍ .05 was considered statistically significant for all statistical tests.
1468
Academic Radiology, Vol 13, No 12, December 2006
Agreement between the two MSCT readers for the detection or exclusion of significant in-stent stenosis and stenoses proximal and distal to the stent was quantified using kappa () analysis in which ⬍ 0.20 indicates poor agreement; ⫽ 0.21– 0.40 indicates fair agreement; ⫽ 0.41– 0.60 indicates moderate agreement; ⫽ 0.61– 0.80 indicates good agreement; and ⫽ 0.81–1.00 indicates very good agreement.
RESULTS MSCT Success Rate and Image Quality MSCT was performed successfully in all 25 patients. According to the preliminary qualitative consensus analysis of the two MSCT readers, the examined stents were depicted with high image quality in 65% (30/46) of cases, moderate image quality in 33% (15/46) of cases, and poor image quality in 2% (1/46) of cases. The single case of poor image quality concerned a small (2.5 mm diameter) stent, which was uninterpretable on MSCT because of extensive metallic streak artifacts. Overall, the artificial narrowing of the in-stent lumen and of the vessel lumen proximal and distal to the stent was markedly less with the B46f convolution kernel than with the B30f convolution kernel. Coronary Angiography Invasive coronary angiography revealed significant in-stent restenosis (⬎50% diameter stenosis) or occlusion in 8/45 (18%) coronary artery stents (six cases of significant stenoses, two cases of stent occlusion). The remaining 37/45 (82%) stents were considered normal or with nonsignificant disease on conventional coronary angiography. Significant stenosis (⬎50% diameter stenosis) occurring within 3 mm proximal and distal to the stent was detected in 4/45 (9%) cases and 3/45 (7%) cases, respectively. The remaining persistent areas were considered to be normal or with nonsignificant disease on conventional coronary angiography. Diagnostic Accuracy of 64SCT In-stent analysis (ⱖ50% diameter restenosis).—Readers 1 and 2 independently determined sensitivity values of 63% and 75%, respectively, and specificity values of 81% and 95%, respectively, for the detection of significant instent stenoses on 64SCT angiography (Table 3). Subsequent consensus reading of all 64SCT images determined a correct final diagnosis for 40 of 45 evaluated stents, for
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Table 3 Sensitivity and Specificity for the Depiction of Significant Proximal, Distal, and In-Stent Stenosis Reader 1
Reader 2
Location of Stenoses
Sensitivity
Specificity
Sensitivity
Specificity
Reader Agreement (Kappa)
In-stent stenoses Proximal stenoses Distal stenoses
63% (5/8) 50% (2/4) 67% (2/3)
81% (30/37) 88% (36/41) 81% (34/42)
75% (6/8) 75% (3/4) 67% (2/3)
95% (35/37) 95% (39/41) 87% (37/42)
0.53 0.65 0.66
Consensus Sensitivity
Specificity
75% (6/8) 75% (3/4) 67% (2/3)
92% (34/37) 95% (39/41) 86% (36/42)
Overall assessment by two readers for the detection of significant (ⱖ50%) proximal, distal and in-stent stenoses in 64-slice computed tomography coronary angiography in comparison to quantitative X-ray coronary angiography.
an overall diagnostic accuracy of 89%. Overall, two falsenegative observations (in-stent lumen narrowing not detected) and three false-positive observations were recorded on the consensus reading, giving values for sensitivity and specificity of 75% (6 of 8 significant in-stent stenoses or occlusions detected) and 92% (34 of 37 correct negative diagnoses), respectively. The sensitivity for detecting significantly diseased but not occluded stents was 67% (4 of 6), whereas the sensitivity for detecting stent occlusions was 100% (2 of 2). The two cases in which restenosis was not detected occurred in stents with small diameters (2.5 mm and 3.0 mm, respectively; correct assessment of stent patency is shown in Fig 1, falsenegative evaluation of an in-stent restenosis is demonstrated in Fig 2). Based on the consensus reading the PPV and NPV for the detection of significant in-stent stenoses were 67% (6 of 9) and 94% (34 of 36), respectively. Based on the total patient population the degree of interreader agreement was considered to be moderate ( ⫽ 0.53; standard error: 0.26; 95% confidence interval: 0.026 –1.04). Analysis of coronary artery lumen proximal to stent (ⱖ50% diameter restenosis).—Of the four significantly diseased proximal stent segments on QCA, readers 1 and 2 correctly detected two and three stenoses, respectively, on 64SCT (Table 3). On consensus reading, three of four proximal stent stenoses were detected on 64SCT with two false-positive observations and one false-negative observation. Based on these findings, the values for sensitivity, specificity and diagnostic accuracy for the detection of significant proximal stent stenosis were 75% (3 of 4), 95% (39 of 41), and 93% (42 of 45), respectively. The case in which peri-stent stenosis was not detected occurred in a stent with a diameter of 2.5 mm. The corresponding PPV and NPV were 60% (3 of 5) and 97% (39 of 40), respectively. The overall reader agreement for the evaluation of proximal stent segments was good ( ⫽
0.65; standard error: 0.13; 95% confidence interval: 0.40 – 0.91). Correct assessment of peri-stent patency is demonstrated in Fig 3. Analysis of coronary artery lumen distal to stent (ⱖ50% diameter restenosis).—Of the three significantly diseased distal stent segments on QCA, readers 1 and 2 each correctly detected two stenoses on 64SCT (Table 3). The case in which persistent stenosis was not detected occurred in a stent with a diameter of 3 mm. Consensus reading determined a correct diagnosis on 64SCT for 38 of 45 distal stent segments, with 6 false-positive observations and 1 false-negative observation. The values for sensitivity, specificity, and overall diagnostic accuracy for the detection of significant distal stent stenoses were therefore 67% (2 of 3), 86% (36 of 42), and 84% (38 of 45), respectively, whereas the PPV and NPV were 25% (2 of 8) and 97% (36 of 37), respectively. The reader agreement for the evaluation of distal stent segments was good ( ⫽ 0.66; standard error: 0.12; 95% confidence interval: 0.42– 0.89).
DISCUSSION Follow-up examinations of patients after coronary artery stent placement is an important clinical issue in daily cardiology routine because of often very high rates of restenosis; restenosis rates of up to 46% have been reported during the first 6 months when non– drug-eluting stents are used (2,3). The early detection of in-stent restenosis is crucial to avoid myocardial ischemia and to improve long-term prognosis: in-stent restenosis has a marked negative impact on the long-term survival of patients treated with coronary artery stents (18). Unfortunately, noninvasive tools have so far shown only low to moderate sensitivity for the detection of in-stent restenosis
1469
RIST ET AL
Academic Radiology, Vol 13, No 12, December 2006
Figure 1. MSCT images and conventional angiography of a 74-year-old male patient after stent placement in the left anterior descending (LAD) artery. Stent patency was confirmed by conventional x-ray angiography of the left coronary system (d). Quantitative coronary angiography for stented coronary artery lumen (LAD) shows no relevant lumen reduction (d, arrow). Applying volume rendering techniques (c, arrow) as well as multiplanar reformats (a, arrow) the patency of the stent was confirmed by 64-slice computed tomography (64SCT). No in-stent plaque or stenoses can be detected and vessel opacification proximal and distal to the stent is normal. Because of the high spatial resolution of 64SCT, even cross-sectional imaging of stents becomes feasible, enabling the differentiation of the in-stent lumen and the stent struts (b, arrow).
(19). A noninvasive alternative for the evaluation of stented coronary arteries would therefore be highly desirable. Coronary MSCT is a robust method for noninvasive imaging of the coronary arteries, and is under investigation as a possible alternative to conventional angiography for dedicated applications (eg, exclusion of coronary artery disease) (20). In our study population of 25 patients referred for 64SCT, only a small percentage (2%; 1 of 46 stents) was inadequate for diagnostic evaluation. The cases in which image quality was considered poor or
1470
moderate typically involved patients fitted with small diameter stents in distal segments of the coronary arteries. The improved image quality in our study compared with previous studies can be ascribed to the improved spatial and temporal resolution achievable on 64SCT compared with 4SCT and 16SCT systems. Specifically, the 64-slice scanner used in this study makes use of a periodic motion of the focal spot in the longitudinal direction (z-flying focal spot) to double the number of simultaneously acquired slices (14). This technical advance permits 64
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Figure 2. Multislice computed tomography datasets and conventional x-ray angiography of 65-yearold male patient after stent placement in the proximal left anterior descending artery. False-negative evaluation of an in-stent restenosis. Conventional x-ray angiography shows significant stenosis (55%) in the stented area (a,b) magnified view of the stented area, arrow. In the corresponding multiplanar reformatted image (c, multiplanar reconstruction, d, magnified view, arrow), no significant restenosis was visible.
overlapping 0.6-mm slices to be acquired per rotation and enables the rotation time of the scanner to be reduced to 330 milliseconds, resulting in a temporal resolution of 83–165 milliseconds depending on the heart rate. The faster scan duration of 64SCT (about 10 –12 seconds) compared with 4SCT and 16SCT systems not only permits a reduction of the volume of contrast medium required, but also minimizes artifacts from breathing motion, thereby improving the robustness of the technique. The improved performance of 64SCT compared with 4SCT and 16SCT also reduces the influence of heart rate on the examination. In this regard, earlier studies to evaluate MSCT coronary angiography were limited by the
presence of a substantial number of subjects with high (⬎65 bpm) heart rates, changing heart rates, or arrhythmias. Other studies enrolled only preselected subjects with low heart rates (⬍65 bpm) or routinely administered beta-blockers to achieve an acceptable image quality (8,21). These limitations of MSCT have largely been overcome with 64SCT; in our study, image quality and the assessment of stents were not significantly affected by high heart rates or arrhythmias. Although partial volume effects and beam-hardening artifacts were present adjacent to some stented vessels depending on the stent type and strut architecture, evaluation of the stent lumen was nevertheless still possible in 98% of cases.
1471
RIST ET AL
Academic Radiology, Vol 13, No 12, December 2006
Figure 3. Multislice computed tomography (MSCT) images and conventional x-ray angiography of a 58-year-old male patient after placement of two sequential stents in the proximal left anterior descending artery. The multiplanar reformat of the MSCT dataset shows a predominant noncalcified plaque before the stent lumen (a, arrowhead); no relevant stenosis between sequential stents can be delineated (a, arrow). Quantitative coronary angiography shows no significant stenosis between proximal stent and distal stent (b, arrow).
The sensitivity and specificity for the detection of significant in-stent disease or occlusion was 75% and 92%, respectively, on consensus reading. Significant (ⱖ50%) in-stent restenosis was noted in four of six cases, whereas in-stent occlusion was correctly identified on 64SCT in two of two cases. A recent study (11) in which 16SCT was used in the assessment of stent patency reported similar values for sensitivity (78%) and specificity (100%) for the detection of significant in-stent stenoses. However, as mentioned previously, a large proportion of the stents evaluated showed insufficient image quality to fully assess stent patency. Another recent study has similarly shown that 16SCT permits the correct diagnosis of in-stent lumen in a large majority (27 of 29) of cases after stent placement, although in this study only the left main arteries with comparatively large diameters were evaluated (22). Similar overall findings to those observed for in-stent disease were noted for the evaluation of peri-stent disease. Comparison of the sensitivity and specificity values for detection of proximal and distal stent stenoses revealed higher values for the former. Although recent phantom (13,23) and ex vivo studies (24) have shown superior stent lumen visibility on 64SCT scanners, even on the most advanced systems, the diagnostic accuracy at distal vessel segments is limited (25–28). In our study, of seven cases of peri-stent restenosis, three stenoses were not correctly identified. Each of these three stents was
1472
located in a small caliber distal coronary artery segment. On the other hand, it should be noted that a contrast medium with a standard iodine concentration (300 mg iodine/mL) was used in this study. Recently, it has been shown that higher concentrations of iodine delivered at the same volume and injection rate produce significantly higher attenuation of the coronary arteries (29,30). Notably, the visualization of smaller coronary arteries is improved with higher concentrations of iodine. It is possible that the three peri-stent stenoses that were not identified in this study would have been identified had a contrast medium with a high concentration of iodine (ⱖ370 mg iodine/mL) been used. Moreover, it is possible the overall results in terms of sensitivity and specificity would have been better. Limitations and Future Perspectives Among the principal limitations of our study is that morphologic evaluation of the MSCT datasets does not allow interpretation of the hemodynamic relevance of instent disease. Another important limitation was that not all datasets were reconstructed with the dedicated B46f convolution kernel. Because artificial narrowing of the in-stent lumen and of the vessel lumen proximal and distal to the stent was less with the B46f convolution kernel than with the B30f convolution kernel, it is possible that the consistent use of the B46f convolution kernel would have permitted improved overall stent visualization and hence better overall results.
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
A final limitation is that the radiation exposure required for the protocol was still relatively high, despite the considerable technical improvements of the 64SCT system. To address this particular limitation, new technical developments for the reduction of radiation dose should be introduced to maintain radiation exposure in an acceptable range. Finally, only a small percentage of the patients enrolled in our study had significant restenoses or stent occlusion. For this reason, our results should be considered as preliminary findings and the data concerning the sensitivity in assessing significant stenoses should be interpreted carefully. Moreover, further work in a larger patient population should be performed. CONCLUSION The temporal and spatial resolution of interventional x-ray coronary angiography is still unmatched. However, the new generation of 64SCT scanners provides a significantly increased spatial and temporal resolution when compared to earlier MSCT systems. Our initial data raise the hope that in a clinical setting these benefits could be advantageous for the reliable, noninvasive diagnosis or exclusion of significant in-stent stenosis and the detection of significant peri-stent disease. Although diagnostic image quality is reproducible and stent patency can be assessed correctly in most cases, the detection of in-stent disease remains a major challenge and depends strongly on stent diameter. Future studies will be needed to confirm whether MSCT will play an important role for patients after coronary interventions and stent placement, and whether MSCT will assist in clinical management and outcomes. REFERENCES 1. American Heart Association. Heart and stroke statistics—2004 update. Dallas, TX: American Heart Association; 2004. 2. Antoniucci D, Valenti R, Santoro GM, et al. Restenosis after coronary stenting in current clinical practice. Am Heart J 1998; 135:510 – 8. 3. Gordon PC, Gibson CM, Cohen DJ, Carrozza JP, Kuntz RE, Baim DS. Mechanisms of restenosis and redilation within coronary stents— quantitative angiographic assessment. J Am Coll Cardiol 1993; 21:1166 –1174. 4. Peterson ED, Cowper PA, DeLong ER, et al. Acute and long-term cost implications of coronary stenting. J Am Coll Cardiol 1999; 33:1610 –1618. 5. Regar E, Serruys PW, Bode C, et al. Angiographic findings of the multicenter Randomized Study With the Sirolimus-Eluting Bx Velocity Balloon-Expandable Stent (RAVEL): sirolimus-eluting stents inhibit restenosis irrespective of the vessel size. Circulation 2002; 106:1949 –1956. 6. Grube E, Silber S, Hauptmann KE, et al. TAXUS I: six- and twelve-month results from a randomized, double-blind trial on a slow-release paclitaxeleluting stent for de novo coronary lesions. Circulation 2003; 107:38 – 42. 7. Ohnesorge B, Flohr T, Becker C, et al. Cardiac imaging by means of electrocardiographically gated multisection spiral CT: initial experience. Radiology 2000; 217:564 –571. 8. Ropers D, Baum U, Pohle K, et al. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003; 107:664 – 666.
9. Maintz D, Grude M, Fallenberg EM, et al. Assessment of coronary arterial stents by multislice-CT angiography. Acta Radiol 2003; 44:597– 603. 10. Mahnken AH, Buecker A, Wildberger JE, et al. Coronary artery stents in multislice computed tomography: in vitro artifact evaluation. Invest Radiol 2004; 39:27–33. 11. Schuijf JD, Bax JJ, Jukema JW, et al. Feasibility of assessment of coronary stent patency using 16-slice computed tomography. Am J Cardiol 2004; 94:427– 430. 12. Gaspar T, Halon DA, Lewis BS, et al. Diagnosis of coronary in-stent restenosis with multidetector row spiral computed tomography. J Am Coll Cardiol 2005; 46:1573–1579. 13. Seifarth H, Ozgun M, Raupach R, et al. 64- versus 16-slice CT angiography for coronary artery stent assessment: in vitro experience. Invest Radiol 2006; 41:22–27. 14. Flohr T, Stierstorfer K, Raupach R, et al. Performance evaluation of a 64-slice CT system with z-flying focal spot. Rofo 2004; 176:1803–1810. 15. Cademartiri F, Nieman K, van der Lugt A, et al. Intravenous contrast material administration at 16-detector row helical CT coronary angiography: test bolus versus bolus-tracking technique. Radiology 2004; 233:817– 823. 16. Jakobs TF, Becker CR, Ohnesorge B, et al. Multislice helical CT of the heart with retrospective ECG gating: reduction of radiation exposure by ECG-controlled tube current modulation. Eur Radiol 2002; 12:1081–1086. 17. Dodge JT, Brown BG, Bolson EL, et al. Intrathoracic spatial location of coronary segments on the normal human heart. Circulation 1988; 78: 1167–1180. 18. Kastrati A, Mehilli J, von Beckerath N, et al. Sirolimus-eluting stent or paclitaxel-eluting stent vs balloon angioplasty for prevention of recurrences in patients with coronary in-stent restenosis: a randomized controlled trial. JAMA 2005; 293:165–171. 19. Galassi AR, Foti R, Azzarelli S, et al. Usefulness of exercise tomographic myocardial perfusion imaging for detection of restenosis after coronary stent implantation. Am J Cardiol 2000; 85:1362–1364. 20. Becker CR, Knez A, Leber A, et al. Detection of coronary artery stenoses with multislice helical CT angiography. J Comput Assist Tomogr 2002; 26:750 –755. 21. Kuettner A, Trabold T, Schroeder S, et al. Noninvasive detection of coronary lesions using 16-detector multislice spiral computed tomography technology: initial clinical results. J Am Coll Cardiol 2004; 44:1230 –1237. 22. Gilard M, Cornily JC, Rioufol G, et al. Noninvasive assessment of left main coronary stent patency with 16-slice computed tomography. Am J Cardiol 2005; 95:110 –112. 23. Maintz D, Seifarth H, Raupach R, et al. 64-slice multidetector coronary CT angiography: in vitro evaluation of 68 different stents. Eur Radiol. 2006 Jul; 16(7):1564 –9. Epub 2006 Mar 2. 2006 Apr; 16(4):818 –26. Epub 2005 Dec 7. 24. Rist C, Nikolaou K, Flohr T, et al. High-resolution ex vivo imaging of coronary artery stents using 64-slice computed tomography—initial experience. Eur Radiol. In press. 25. Nikolaou K, Knez A, Rist C, et al. 64-slice computed tomography in the diagnosis of ischemic heart disease. Am J Roentgenol. 2006 Jul; 187(1): 111–7. 26. Mollet NR, Cademartiri F, van Mieghem CA, et al. High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation 2005; 112: 2318 –2323. 27. Leschka S, Alkadhi H, Plass A, et al. Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J 2005; 26:1482–1487. 28. Nikolaou K, Flohr T, Knez A, et al. Advances in cardiac CT imaging: 64-slice scanner. Int J Cardiovasc Imaging 2004; 20:535–540. 29. Cademartiri F, de Monye C, Pugliese F, et al. High iodine concentration contrast material for noninvasive multislice computed tomography coronary angiography: iopromide 370 versus iomeprol 400. Invest Radiol 2006; 41:349 –353. 30. Cademartiri F, Mollet NR, van der Lugt A, et al. Intravenous contrast material administration at helical 16-detector row CT coronary angiography: effect of iodine concentration on vascular attenuation. Radiology 2005; 236:661– 665.
1473
Rationale and Objectives. Restenosis remains a major limitation of coronary catheter– based stent placement. Therefore, a reliable noninvasive diagnostic method for the evaluation of stented coronary arteries would be highly desirable. Our aim was to evaluate the diagnostic accuracy of high-resolution 64-slice computed tomography (64SCT) in a pilot study for the assessment of the lumen of coronary artery stents. Materials and Methods. Twenty-five patients underwent 64SCT of the coronary arteries and quantitative x-ray coronary angiography (QCA) after coronary artery stent placement. 64SCT coronary angiography was performed with the following parameters: spatial resolution ⫽ 0.4 ⫻ 0.4 ⫻ 0.4 mm; temporal resolution ⫽ 83–165 milliseconds; contrast agent ⫽ 80 mL at a flow rate of 5 mL/second; retrospective electrocardiogram gating. The 64SCT scans were evaluated for image quality and for the presence of significant in-stent and peri-stent (proximal and distal) stenoses. Determinations were made of the sensitivity, specificity, diagnostic accuracy, and positive and negative predictive values (PPV and NPV) of 64SCT for the detection or exclusion of stenoses. Results. A total of 46 stents were evaluated, of which 45 (98%) were of diagnostic image quality. Significant in-stent restenosis or occlusion was detected on QCA in 8/45 cases (ⱖ50% stenosis ⫽ 6; occlusion ⫽ 2). The sensitivity, specificity, accuracy, PPV, and NPV of 64SCT for the detection of significant in-stent disease was 75%, 92%, 89%, 67%, and 94%, respectively. Both occluded coronary artery stents were correctly identified. The sensitivity, specificity, and accuracy values of 64SCT for the detection of significant proximal peri-stent stenoses were 75%, 95%, and 93%, respectively, whereas the values for detection of significant distal peri-stent stenoses were 67%, 85%, and 84%, respectively. Conclusion. The high spatial and temporal resolution of 64SCT may permit improved assessment of stent occlusion and peri-stent disease, although detection of in-stent stenosis remains difficult. Key Words. Coronary artery disease; multislice computed tomography; atherosclerosis; coronary artery stent. ©
AUR, 2006
Coronary artery stenting is the predominant method of nonsurgical myocardial revascularization; an estimated 475,000 coronary artery stents were implanted in the United States alone in 2001 (1). Unfortunately, de novo coronary lesions
Acad Radiol 2006; 13:1465–1473 1
From the Department of Clinical Radiology (C.R., K.N., B.J.W., T.J., M.F.R., C.R.B.) and Department of Cardiology (F.v.Z., A.K., A.W.L.), University Hospitals – Grosshadern Ludwig-Maximilians University, Marchioninistr. 15, 81377 Munich, Germany, and Worldwide Medical Affairs, Bracco Imaging SpA, Milan, Italy (M.A.K.). Received June 29, 2006; accepted September 8, 2006. C.R. and F.v.Z. have contributed equally to this article. Address correspondence to: C.R. e-mail: [email protected]
© AUR, 2006 doi:10.1016/j.acra.2006.09.044
and restenosis are major problems in a considerable proportion of patients; a 6-month restenosis rate ranging from 11% to 46% has been reported if non– drug-eluting stents are used (2,3). Patients with restenosis often require repeat interventional procedures that result in a reduced overall quality of life and increased costs to the health care system (4). Initial clinical experience with drug-eluting coronary stents has shown excellent results, with 0% restenosis at 4-month, 6-month, and 12-month follow-up examinations (5,6). Nevertheless, despite these recent findings, restenosis remains the major limitation of coronary catheter-based intervention. A reliable noninvasive diagnostic method for the evaluation of stented coronary arteries would therefore be highly desirable.
1465
RIST ET AL
The introduction of multislice computed tomography (MSCT) technology combined with fast rotation time and retrospective electrocardiogram gating permits the entire heart to be examined in a single breath hold, while maintaining high submillimeter isotropic spatial resolution (7). However, whereas the important role of 16-slice CT systems (16SCT) in MSCT coronary angiography for the diagnosis of coronary artery disease is well-established (8), the potential benefits for evaluation of coronary artery stents has still to be established. Early studies using a four-slice CT scanner revealed frequent false-positive diagnoses of high-grade in-stent stenoses because of artificial thickening of the stent struts (9). More recent in vitro experiments on a 16SCT system to assess different stent types in terms of stent lumen visualization revealed that visualization was largely dependent on the metallic composition and diameter of the particular stent (10). Promising results for the visualization of stent lumen have subsequently been obtained with 16SCT and with 40/64-slice (64SCT) for the detection of significant in-stent stenosis (11–13). The purpose of our study was to investigate the potential of MSCT coronary angiography using a 64SCT system for the assessment of stented coronary arteries. Specifically, our study was aimed at evaluating the sensitivity, specificity, and diagnostic accuracy of 64SCT coronary angiography for the detection of significant instent and peri-stent steno-occlusive disease.
MATERIAL AND METHODS Patient Population The study population comprised 25 consecutive patients (23 male, 2 female; mean age 59.4 ⫾ 12 years; range 40 – 83 years). Patients were included over a period of 4 months. Eligible patients had each undergone percutaneous transluminal coronary angioplasty with stent placement. After this interventional procedure, all patients underwent a second conventional coronary angiography and a comparative MSCT angiography of the coronary arteries. The mean time interval between MSCT and this second conventional angiography was 5 ⫾ 2 days (range 1–9 days), whereas MSCT was performed at a mean interval of 22 ⫾ 41.8 days (range 1–182 days) after coronary stent placement. For the majority of our study population (n ⫽ 17), the indication for conventional angiography after stent placement was the suspicion of in-stent restenosis because of new or changing clinical symptoms.
1466
Academic Radiology, Vol 13, No 12, December 2006
For the remaining patients (n ⫽ 8), the indication for catheter-based coronary angiography was an Institutional Review Board–approved clinical study to evaluate drugeluting and drug-coated stents versus non– drug-eluting stents over a 3-month period. Exclusion criteria were 1) atrial fibrillation and other severe arrhythmias, 2) severe renal failure, 3) hyperthyroidism, 4) hypersensitivity to iodine-containing contrast agent, 5) pregnancy, and 6) unstable clinical conditions. The mean heart rate during the MSCT examinations was 62.2 ⫾ 8 bpm (range 49 –79 bpm). Beta-blocker treatment (Metoprolol 5–10 mg) was administered intravenously to patients (n ⫽ 14) with heart rates ⬎70 bpm shortly before the MSCT examination. Metoprolol was not administered to patients with known hypersensitivity or contraindications for beta-blockers. All patients gave informed consent to participate in the study. The study protocol was approved by the Ethics Committee of the hospital. Stent Characteristics A total of 46 stents were implanted in the 25 patients included in the study. Nine different stent types were employed; the material and design of these stents are summarized in Table 1. The majority of the stents used (27/46) were made of surgical stainless steel (316 L). The remaining 19 stents comprised 6 made of chrome-cobalt composition, 7 that were paclitaxel-eluting/coated stents, and 6 that were rapamycin-eluting/coated stents. The diameter of the implanted stents ranged from 2.5 mm to 4.0 mm. The site of stent implantation was the right coronary artery in 17 cases, the left main in 1 case, the left anterior descending in 20 cases, and the left circumflex in 8 cases (Table 2). MSCT Scanning Protocol MSCT was performed using a new-generation 64SCT system (Somatom Sensation 64, Siemens Medical Solutions, Forchheim, Germany) working at an increased rotation rate and with application of z-axis flying focal spot technology to achieve an effective slice-thickness of 0.6 mm (14). A contrast medium bolus of 80 mL (Iomeron 300, Bracco Imaging SpA, Milan, Italy) was injected into the antecubital vein at a flow rate of 5 mL/second. Automated bolus tracking software was employed to time the start of the scan to the arrival of the contrast medium bolus; each scan began automatically 6 seconds after the contrast density in the ascending aorta reached a predefined threshold of 100 Hounsfield units (15). Images
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Table 1 Properties of Balloon-Expandable Stents (n ⴝ 46) and Commercial Name of Stents, Manufacturer, Material, Length, and Nominal Diameter of the Examined Stents
Name Paclitaxel MAC Arthos Bx-Sonic Cypher select Arthos Pico Flex Master Bx-Velocity Taxus
Manufacturer
Material
Cook Mönchengladbach, Germany AMG Raesfeld-Erle, Germany AMG Raesfeld-Erle, Germany Cordis Langenfeld, Germany Cordis Langenfeld, Germany
Stainless steel 316 L Paclitaxel Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Rapamycin Chrome-Cobalt composition Stainless steel 316 L Stainless steel 316 L Stainless steel 316 L Paclitaxel
AMG Raesfeld-Erle, Germany Jomed, Haan, Germany Cordis Langenfeld, Germany Boston Scientific Ratingen, Germany
Table 2 Distribution of Implanted Coronary Artery Stents According to the AHA-15 Segment Coronary Artery Model
RCA
LM LAD
Rd1 Rd2 LCx
Total
Segment
Number of Stents
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
5 7 5 0 1 7 11 0 2 0 6 1 1 0 0 46
AHA, American Heart Association; RCA, right coronary artery; LM, left main artery; LAD, left anterior descending artery; Rd1, first diagonal branch; Rd2, second diagonal branch; LCx, left circumflex artery.
were acquired during an inspiratory breath-hold; the entire volume of the heart was covered in about 10 seconds. The electrocardiogram signal of the patient was recorded simultaneously during the examination to enable retrospective gating of the data. The gantry rotation time was 330 milliseconds, the tube current was 850 mAs using dose modulation (16), and the tube voltage was 120 kV.
Length (mm)
Strut Dimensions (mm)
Diameter (mm)
Number of Implanted Stents
16
0.125
2.5, 3.5
3
9, 13 33 18 33
0.125 0.125 0.14 0.14
3.5, 4.0 3 3 3.5
6 1 2 6
14, 19 9 8, 13 16
0.065–0.075 0.09 0.14 n/a
2.5, 3.0 3.0 3.0 3.0
6 13 5 4
The temporal resolution was 83–165 milliseconds, the detector collimation was 64 ⫻ 0.6 mm, and the spatial resolution was 0.4 ⫻ 0.4 ⫻ 0.4 mm. For optimal, motion-free image quality, datasets were reconstructed in mid-diastole (mean interval 614 ⫾ 175 milliseconds after the R-wave). Based on these protocol parameters the estimated radiation dose was approximately 8 –10 mSv. Most datasets were reconstructed using a medium smooth body convolution kernel (B30f) as well as a sharp heart view (B46f) convolution kernel. The B46f kernel was used in datasets for 19/25 patients (34/45 stents). Conventional X-Ray Coronary Angiography Both conventional x-ray coronary angiography and balloon angioplasty with stenting procedures were performed according to standard techniques. Coronary angiograms were evaluated quantitatively (QuantCor QCA, Siemens Medical Systems) by an independent investigator who was fully blinded to the findings of MSCT angiography. After visual assessment, quantitative x-ray coronary angiography (QCA) was performed of all stented vessels and proximal and distal coronary artery lumina. The mean diameter reduction of these lesions in two projections was determined and the segmental location of the stenoses according to the 15-segment AHA (American Heart Association model) was recorded (17). MSCT Image Analysis All MSCT datasets were transferred to a commercially available postprocessing workstation (Leonardo, Siemens)
1467
RIST ET AL
and analyzed by two independent experienced readers who were blinded to the results of x-ray coronary angiography. Images were evaluated initially for diagnostic quality by both readers in consensus. A 3-point scale was used: 1 ⫽ high image quality (optimal depiction of the coronary artery and the stented area); 2 ⫽ moderate image quality (sufficient depiction of coronary artery and coronary artery stent for diagnostic purposes, but minor motion artifacts or reduced signal-to-noise ratio); and 3 ⫽ poor or uninterpretable image quality (data not adequate for diagnostic assessment of stent lumina or proximal and distal stent areas). Images considered to be of high or moderate diagnostic quality were then evaluated further by the two readers independently. Evaluation was performed of the original axial source images and of postprocessed curved multiplanar reformations, maximum intensity projection reconstructions, and volume rendered three-dimensional reconstructions. Each coronary artery stent segment was independently evaluated according to a four-point scale, in which 1 ⫽ patent stent lumen, no in-stent disease; 2 ⫽ stent lumen patent, but in-stent restenosis of ⬍50% of the vessel lumen diameter due to atherosclerotic vessel wall changes or intimal hyperplasia; 3 ⫽ significant in-stent restenosis of ⱖ50% of the vessel lumen diameter; and 4 ⫽ stent occlusion. A stent lumen was considered to be patent if contrast medium was detected inside the stent lumen and if peripheral enhancement was present. Finally, independent evaluation was also performed to determine the presence/absence of stenosis 3 mm proximal and distal to the stent (peri-stent disease). Narrowing of the luminal diameter by ⱖ50% was considered to indicate significant proximal or distal stent stenosis. After the independent reading, a consensus reading was performed to obtain a final MSCT diagnosis. Statistical Analysis Statistical analysis was performed using two Windowsbased software products [MedCalc, Version 7.0.0.2, 2002 (Mariakerke, Belgium); SPSS 12.0.1, 2003 (Chicago, Illinois)]. Continuous variables were presented as mean ⫾ standard deviation. The value of 64SCT for the detection of significant coronary artery stent restenosis was determined against conventional angiography as the standard of reference. The sensitivity, specificity, diagnostic accuracy, and positive and negative predictive values (PPV and NPV, respectively) were calculated. A significance level of P ⬍ .05 was considered statistically significant for all statistical tests.
1468
Academic Radiology, Vol 13, No 12, December 2006
Agreement between the two MSCT readers for the detection or exclusion of significant in-stent stenosis and stenoses proximal and distal to the stent was quantified using kappa () analysis in which ⬍ 0.20 indicates poor agreement; ⫽ 0.21– 0.40 indicates fair agreement; ⫽ 0.41– 0.60 indicates moderate agreement; ⫽ 0.61– 0.80 indicates good agreement; and ⫽ 0.81–1.00 indicates very good agreement.
RESULTS MSCT Success Rate and Image Quality MSCT was performed successfully in all 25 patients. According to the preliminary qualitative consensus analysis of the two MSCT readers, the examined stents were depicted with high image quality in 65% (30/46) of cases, moderate image quality in 33% (15/46) of cases, and poor image quality in 2% (1/46) of cases. The single case of poor image quality concerned a small (2.5 mm diameter) stent, which was uninterpretable on MSCT because of extensive metallic streak artifacts. Overall, the artificial narrowing of the in-stent lumen and of the vessel lumen proximal and distal to the stent was markedly less with the B46f convolution kernel than with the B30f convolution kernel. Coronary Angiography Invasive coronary angiography revealed significant in-stent restenosis (⬎50% diameter stenosis) or occlusion in 8/45 (18%) coronary artery stents (six cases of significant stenoses, two cases of stent occlusion). The remaining 37/45 (82%) stents were considered normal or with nonsignificant disease on conventional coronary angiography. Significant stenosis (⬎50% diameter stenosis) occurring within 3 mm proximal and distal to the stent was detected in 4/45 (9%) cases and 3/45 (7%) cases, respectively. The remaining persistent areas were considered to be normal or with nonsignificant disease on conventional coronary angiography. Diagnostic Accuracy of 64SCT In-stent analysis (ⱖ50% diameter restenosis).—Readers 1 and 2 independently determined sensitivity values of 63% and 75%, respectively, and specificity values of 81% and 95%, respectively, for the detection of significant instent stenoses on 64SCT angiography (Table 3). Subsequent consensus reading of all 64SCT images determined a correct final diagnosis for 40 of 45 evaluated stents, for
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Table 3 Sensitivity and Specificity for the Depiction of Significant Proximal, Distal, and In-Stent Stenosis Reader 1
Reader 2
Location of Stenoses
Sensitivity
Specificity
Sensitivity
Specificity
Reader Agreement (Kappa)
In-stent stenoses Proximal stenoses Distal stenoses
63% (5/8) 50% (2/4) 67% (2/3)
81% (30/37) 88% (36/41) 81% (34/42)
75% (6/8) 75% (3/4) 67% (2/3)
95% (35/37) 95% (39/41) 87% (37/42)
0.53 0.65 0.66
Consensus Sensitivity
Specificity
75% (6/8) 75% (3/4) 67% (2/3)
92% (34/37) 95% (39/41) 86% (36/42)
Overall assessment by two readers for the detection of significant (ⱖ50%) proximal, distal and in-stent stenoses in 64-slice computed tomography coronary angiography in comparison to quantitative X-ray coronary angiography.
an overall diagnostic accuracy of 89%. Overall, two falsenegative observations (in-stent lumen narrowing not detected) and three false-positive observations were recorded on the consensus reading, giving values for sensitivity and specificity of 75% (6 of 8 significant in-stent stenoses or occlusions detected) and 92% (34 of 37 correct negative diagnoses), respectively. The sensitivity for detecting significantly diseased but not occluded stents was 67% (4 of 6), whereas the sensitivity for detecting stent occlusions was 100% (2 of 2). The two cases in which restenosis was not detected occurred in stents with small diameters (2.5 mm and 3.0 mm, respectively; correct assessment of stent patency is shown in Fig 1, falsenegative evaluation of an in-stent restenosis is demonstrated in Fig 2). Based on the consensus reading the PPV and NPV for the detection of significant in-stent stenoses were 67% (6 of 9) and 94% (34 of 36), respectively. Based on the total patient population the degree of interreader agreement was considered to be moderate ( ⫽ 0.53; standard error: 0.26; 95% confidence interval: 0.026 –1.04). Analysis of coronary artery lumen proximal to stent (ⱖ50% diameter restenosis).—Of the four significantly diseased proximal stent segments on QCA, readers 1 and 2 correctly detected two and three stenoses, respectively, on 64SCT (Table 3). On consensus reading, three of four proximal stent stenoses were detected on 64SCT with two false-positive observations and one false-negative observation. Based on these findings, the values for sensitivity, specificity and diagnostic accuracy for the detection of significant proximal stent stenosis were 75% (3 of 4), 95% (39 of 41), and 93% (42 of 45), respectively. The case in which peri-stent stenosis was not detected occurred in a stent with a diameter of 2.5 mm. The corresponding PPV and NPV were 60% (3 of 5) and 97% (39 of 40), respectively. The overall reader agreement for the evaluation of proximal stent segments was good ( ⫽
0.65; standard error: 0.13; 95% confidence interval: 0.40 – 0.91). Correct assessment of peri-stent patency is demonstrated in Fig 3. Analysis of coronary artery lumen distal to stent (ⱖ50% diameter restenosis).—Of the three significantly diseased distal stent segments on QCA, readers 1 and 2 each correctly detected two stenoses on 64SCT (Table 3). The case in which persistent stenosis was not detected occurred in a stent with a diameter of 3 mm. Consensus reading determined a correct diagnosis on 64SCT for 38 of 45 distal stent segments, with 6 false-positive observations and 1 false-negative observation. The values for sensitivity, specificity, and overall diagnostic accuracy for the detection of significant distal stent stenoses were therefore 67% (2 of 3), 86% (36 of 42), and 84% (38 of 45), respectively, whereas the PPV and NPV were 25% (2 of 8) and 97% (36 of 37), respectively. The reader agreement for the evaluation of distal stent segments was good ( ⫽ 0.66; standard error: 0.12; 95% confidence interval: 0.42– 0.89).
DISCUSSION Follow-up examinations of patients after coronary artery stent placement is an important clinical issue in daily cardiology routine because of often very high rates of restenosis; restenosis rates of up to 46% have been reported during the first 6 months when non– drug-eluting stents are used (2,3). The early detection of in-stent restenosis is crucial to avoid myocardial ischemia and to improve long-term prognosis: in-stent restenosis has a marked negative impact on the long-term survival of patients treated with coronary artery stents (18). Unfortunately, noninvasive tools have so far shown only low to moderate sensitivity for the detection of in-stent restenosis
1469
RIST ET AL
Academic Radiology, Vol 13, No 12, December 2006
Figure 1. MSCT images and conventional angiography of a 74-year-old male patient after stent placement in the left anterior descending (LAD) artery. Stent patency was confirmed by conventional x-ray angiography of the left coronary system (d). Quantitative coronary angiography for stented coronary artery lumen (LAD) shows no relevant lumen reduction (d, arrow). Applying volume rendering techniques (c, arrow) as well as multiplanar reformats (a, arrow) the patency of the stent was confirmed by 64-slice computed tomography (64SCT). No in-stent plaque or stenoses can be detected and vessel opacification proximal and distal to the stent is normal. Because of the high spatial resolution of 64SCT, even cross-sectional imaging of stents becomes feasible, enabling the differentiation of the in-stent lumen and the stent struts (b, arrow).
(19). A noninvasive alternative for the evaluation of stented coronary arteries would therefore be highly desirable. Coronary MSCT is a robust method for noninvasive imaging of the coronary arteries, and is under investigation as a possible alternative to conventional angiography for dedicated applications (eg, exclusion of coronary artery disease) (20). In our study population of 25 patients referred for 64SCT, only a small percentage (2%; 1 of 46 stents) was inadequate for diagnostic evaluation. The cases in which image quality was considered poor or
1470
moderate typically involved patients fitted with small diameter stents in distal segments of the coronary arteries. The improved image quality in our study compared with previous studies can be ascribed to the improved spatial and temporal resolution achievable on 64SCT compared with 4SCT and 16SCT systems. Specifically, the 64-slice scanner used in this study makes use of a periodic motion of the focal spot in the longitudinal direction (z-flying focal spot) to double the number of simultaneously acquired slices (14). This technical advance permits 64
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
Figure 2. Multislice computed tomography datasets and conventional x-ray angiography of 65-yearold male patient after stent placement in the proximal left anterior descending artery. False-negative evaluation of an in-stent restenosis. Conventional x-ray angiography shows significant stenosis (55%) in the stented area (a,b) magnified view of the stented area, arrow. In the corresponding multiplanar reformatted image (c, multiplanar reconstruction, d, magnified view, arrow), no significant restenosis was visible.
overlapping 0.6-mm slices to be acquired per rotation and enables the rotation time of the scanner to be reduced to 330 milliseconds, resulting in a temporal resolution of 83–165 milliseconds depending on the heart rate. The faster scan duration of 64SCT (about 10 –12 seconds) compared with 4SCT and 16SCT systems not only permits a reduction of the volume of contrast medium required, but also minimizes artifacts from breathing motion, thereby improving the robustness of the technique. The improved performance of 64SCT compared with 4SCT and 16SCT also reduces the influence of heart rate on the examination. In this regard, earlier studies to evaluate MSCT coronary angiography were limited by the
presence of a substantial number of subjects with high (⬎65 bpm) heart rates, changing heart rates, or arrhythmias. Other studies enrolled only preselected subjects with low heart rates (⬍65 bpm) or routinely administered beta-blockers to achieve an acceptable image quality (8,21). These limitations of MSCT have largely been overcome with 64SCT; in our study, image quality and the assessment of stents were not significantly affected by high heart rates or arrhythmias. Although partial volume effects and beam-hardening artifacts were present adjacent to some stented vessels depending on the stent type and strut architecture, evaluation of the stent lumen was nevertheless still possible in 98% of cases.
1471
RIST ET AL
Academic Radiology, Vol 13, No 12, December 2006
Figure 3. Multislice computed tomography (MSCT) images and conventional x-ray angiography of a 58-year-old male patient after placement of two sequential stents in the proximal left anterior descending artery. The multiplanar reformat of the MSCT dataset shows a predominant noncalcified plaque before the stent lumen (a, arrowhead); no relevant stenosis between sequential stents can be delineated (a, arrow). Quantitative coronary angiography shows no significant stenosis between proximal stent and distal stent (b, arrow).
The sensitivity and specificity for the detection of significant in-stent disease or occlusion was 75% and 92%, respectively, on consensus reading. Significant (ⱖ50%) in-stent restenosis was noted in four of six cases, whereas in-stent occlusion was correctly identified on 64SCT in two of two cases. A recent study (11) in which 16SCT was used in the assessment of stent patency reported similar values for sensitivity (78%) and specificity (100%) for the detection of significant in-stent stenoses. However, as mentioned previously, a large proportion of the stents evaluated showed insufficient image quality to fully assess stent patency. Another recent study has similarly shown that 16SCT permits the correct diagnosis of in-stent lumen in a large majority (27 of 29) of cases after stent placement, although in this study only the left main arteries with comparatively large diameters were evaluated (22). Similar overall findings to those observed for in-stent disease were noted for the evaluation of peri-stent disease. Comparison of the sensitivity and specificity values for detection of proximal and distal stent stenoses revealed higher values for the former. Although recent phantom (13,23) and ex vivo studies (24) have shown superior stent lumen visibility on 64SCT scanners, even on the most advanced systems, the diagnostic accuracy at distal vessel segments is limited (25–28). In our study, of seven cases of peri-stent restenosis, three stenoses were not correctly identified. Each of these three stents was
1472
located in a small caliber distal coronary artery segment. On the other hand, it should be noted that a contrast medium with a standard iodine concentration (300 mg iodine/mL) was used in this study. Recently, it has been shown that higher concentrations of iodine delivered at the same volume and injection rate produce significantly higher attenuation of the coronary arteries (29,30). Notably, the visualization of smaller coronary arteries is improved with higher concentrations of iodine. It is possible that the three peri-stent stenoses that were not identified in this study would have been identified had a contrast medium with a high concentration of iodine (ⱖ370 mg iodine/mL) been used. Moreover, it is possible the overall results in terms of sensitivity and specificity would have been better. Limitations and Future Perspectives Among the principal limitations of our study is that morphologic evaluation of the MSCT datasets does not allow interpretation of the hemodynamic relevance of instent disease. Another important limitation was that not all datasets were reconstructed with the dedicated B46f convolution kernel. Because artificial narrowing of the in-stent lumen and of the vessel lumen proximal and distal to the stent was less with the B46f convolution kernel than with the B30f convolution kernel, it is possible that the consistent use of the B46f convolution kernel would have permitted improved overall stent visualization and hence better overall results.
Academic Radiology, Vol 13, No 12, December 2006
CORONARY ARTERY STENT ASSESSMENT USING 64-SLICE CT
A final limitation is that the radiation exposure required for the protocol was still relatively high, despite the considerable technical improvements of the 64SCT system. To address this particular limitation, new technical developments for the reduction of radiation dose should be introduced to maintain radiation exposure in an acceptable range. Finally, only a small percentage of the patients enrolled in our study had significant restenoses or stent occlusion. For this reason, our results should be considered as preliminary findings and the data concerning the sensitivity in assessing significant stenoses should be interpreted carefully. Moreover, further work in a larger patient population should be performed. CONCLUSION The temporal and spatial resolution of interventional x-ray coronary angiography is still unmatched. However, the new generation of 64SCT scanners provides a significantly increased spatial and temporal resolution when compared to earlier MSCT systems. Our initial data raise the hope that in a clinical setting these benefits could be advantageous for the reliable, noninvasive diagnosis or exclusion of significant in-stent stenosis and the detection of significant peri-stent disease. Although diagnostic image quality is reproducible and stent patency can be assessed correctly in most cases, the detection of in-stent disease remains a major challenge and depends strongly on stent diameter. Future studies will be needed to confirm whether MSCT will play an important role for patients after coronary interventions and stent placement, and whether MSCT will assist in clinical management and outcomes. REFERENCES 1. American Heart Association. Heart and stroke statistics—2004 update. Dallas, TX: American Heart Association; 2004. 2. Antoniucci D, Valenti R, Santoro GM, et al. Restenosis after coronary stenting in current clinical practice. Am Heart J 1998; 135:510 – 8. 3. Gordon PC, Gibson CM, Cohen DJ, Carrozza JP, Kuntz RE, Baim DS. Mechanisms of restenosis and redilation within coronary stents— quantitative angiographic assessment. J Am Coll Cardiol 1993; 21:1166 –1174. 4. Peterson ED, Cowper PA, DeLong ER, et al. Acute and long-term cost implications of coronary stenting. J Am Coll Cardiol 1999; 33:1610 –1618. 5. Regar E, Serruys PW, Bode C, et al. Angiographic findings of the multicenter Randomized Study With the Sirolimus-Eluting Bx Velocity Balloon-Expandable Stent (RAVEL): sirolimus-eluting stents inhibit restenosis irrespective of the vessel size. Circulation 2002; 106:1949 –1956. 6. Grube E, Silber S, Hauptmann KE, et al. TAXUS I: six- and twelve-month results from a randomized, double-blind trial on a slow-release paclitaxeleluting stent for de novo coronary lesions. Circulation 2003; 107:38 – 42. 7. Ohnesorge B, Flohr T, Becker C, et al. Cardiac imaging by means of electrocardiographically gated multisection spiral CT: initial experience. Radiology 2000; 217:564 –571. 8. Ropers D, Baum U, Pohle K, et al. Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003; 107:664 – 666.
9. Maintz D, Grude M, Fallenberg EM, et al. Assessment of coronary arterial stents by multislice-CT angiography. Acta Radiol 2003; 44:597– 603. 10. Mahnken AH, Buecker A, Wildberger JE, et al. Coronary artery stents in multislice computed tomography: in vitro artifact evaluation. Invest Radiol 2004; 39:27–33. 11. Schuijf JD, Bax JJ, Jukema JW, et al. Feasibility of assessment of coronary stent patency using 16-slice computed tomography. Am J Cardiol 2004; 94:427– 430. 12. Gaspar T, Halon DA, Lewis BS, et al. Diagnosis of coronary in-stent restenosis with multidetector row spiral computed tomography. J Am Coll Cardiol 2005; 46:1573–1579. 13. Seifarth H, Ozgun M, Raupach R, et al. 64- versus 16-slice CT angiography for coronary artery stent assessment: in vitro experience. Invest Radiol 2006; 41:22–27. 14. Flohr T, Stierstorfer K, Raupach R, et al. Performance evaluation of a 64-slice CT system with z-flying focal spot. Rofo 2004; 176:1803–1810. 15. Cademartiri F, Nieman K, van der Lugt A, et al. Intravenous contrast material administration at 16-detector row helical CT coronary angiography: test bolus versus bolus-tracking technique. Radiology 2004; 233:817– 823. 16. Jakobs TF, Becker CR, Ohnesorge B, et al. Multislice helical CT of the heart with retrospective ECG gating: reduction of radiation exposure by ECG-controlled tube current modulation. Eur Radiol 2002; 12:1081–1086. 17. Dodge JT, Brown BG, Bolson EL, et al. Intrathoracic spatial location of coronary segments on the normal human heart. Circulation 1988; 78: 1167–1180. 18. Kastrati A, Mehilli J, von Beckerath N, et al. Sirolimus-eluting stent or paclitaxel-eluting stent vs balloon angioplasty for prevention of recurrences in patients with coronary in-stent restenosis: a randomized controlled trial. JAMA 2005; 293:165–171. 19. Galassi AR, Foti R, Azzarelli S, et al. Usefulness of exercise tomographic myocardial perfusion imaging for detection of restenosis after coronary stent implantation. Am J Cardiol 2000; 85:1362–1364. 20. Becker CR, Knez A, Leber A, et al. Detection of coronary artery stenoses with multislice helical CT angiography. J Comput Assist Tomogr 2002; 26:750 –755. 21. Kuettner A, Trabold T, Schroeder S, et al. Noninvasive detection of coronary lesions using 16-detector multislice spiral computed tomography technology: initial clinical results. J Am Coll Cardiol 2004; 44:1230 –1237. 22. Gilard M, Cornily JC, Rioufol G, et al. Noninvasive assessment of left main coronary stent patency with 16-slice computed tomography. Am J Cardiol 2005; 95:110 –112. 23. Maintz D, Seifarth H, Raupach R, et al. 64-slice multidetector coronary CT angiography: in vitro evaluation of 68 different stents. Eur Radiol. 2006 Jul; 16(7):1564 –9. Epub 2006 Mar 2. 2006 Apr; 16(4):818 –26. Epub 2005 Dec 7. 24. Rist C, Nikolaou K, Flohr T, et al. High-resolution ex vivo imaging of coronary artery stents using 64-slice computed tomography—initial experience. Eur Radiol. In press. 25. Nikolaou K, Knez A, Rist C, et al. 64-slice computed tomography in the diagnosis of ischemic heart disease. Am J Roentgenol. 2006 Jul; 187(1): 111–7. 26. Mollet NR, Cademartiri F, van Mieghem CA, et al. High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation 2005; 112: 2318 –2323. 27. Leschka S, Alkadhi H, Plass A, et al. Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J 2005; 26:1482–1487. 28. Nikolaou K, Flohr T, Knez A, et al. Advances in cardiac CT imaging: 64-slice scanner. Int J Cardiovasc Imaging 2004; 20:535–540. 29. Cademartiri F, de Monye C, Pugliese F, et al. High iodine concentration contrast material for noninvasive multislice computed tomography coronary angiography: iopromide 370 versus iomeprol 400. Invest Radiol 2006; 41:349 –353. 30. Cademartiri F, Mollet NR, van der Lugt A, et al. Intravenous contrast material administration at helical 16-detector row CT coronary angiography: effect of iodine concentration on vascular attenuation. Radiology 2005; 236:661– 665.
1473