Quantification and Detectability of In-Stent Stenosis with CT Angiography and MR Angiography in Arterial Stents In Vitro
Abstract
OBJECTIVE. The purpose of this study was to compare CT angiography (CTA) and MR angiography (MRA) for the detectability of 75% and 95% stenoses in phantoms using six different stents.
MATERIALS AND METHODS. Six different stents (Expander, Hemobahn, SelfX, Smart, Symphony, and Wallstent) were inserted into tubes filled with contrast agent (ioversol or gadoteric acid). To mimic stenoses of 75% and 95% of the patent lumen, 8-mm-diameter nylon cylinders were bored in the central axis (2 mm and 4 mm, respectively) and placed into the stent lumen. Intensity profiles across stenoses on 2-mm coronal reformatted sections of CTA or MRA were compared, and the detectability of the residual lumen was assessed using a subjective score.
RESULTS. CTA showed relative in-stent signal attenuation for the in-stent stenoses of the tested stents ranging from 75% to 100% of the signal intensity of the control. SelfX and Symphony showed further shading of the residual lumen due to beam-hardening artifacts. Overestimation of stenosis was associated with low-grade stenoses in which the border of the lumen was closer to the stent struts. MRA showed relative in-stent signal attenuation of the in-stent stenoses ranging from 30% to 100% of the signal intensity of the control. Strut thickness tended to correlate with higher attenuation at CT.
CONCLUSION. CTA may be more suitable for differentiation between 95% stenosis and occlusion; MRA has higher sensitivity in detecting 75% stenoses. Strut thickness and mesh size did not prove to be significant predictors for signal attenuation or overall image quality.
Introduction
With the advent of MDCT and moving-table MRI acquisition, CT angiography and MR angiography are approaching the diagnostic accuracy of conventional angiography [1–5]. CT and MR angiography can be performed on an outpatient basis and are potentially more cost-effective than invasive methods. A further advantage is the lower renal burden resulting from the lower nephrotoxicity of MR angiographic contrast material [6–8]. These techniques are gradually replacing conventional subtraction angiography in the diagnostic assessment of vascular status and the planning of interventional procedures. In the presence of metallic stents, however, the diagnostic accuracy of CT and MR angiography is impaired [9, 10]. CT angiography is prone to beam-hardening effects, and MR angiography suffers from susceptibility artifacts in the presence of metallic stents [7].
Assessment after iliofemoral stent placement ideally must address two main issues. First, hemodynamically relevant stenosis must be accurately detected and differentiated from low-grade stenosis [11]. Second, in cases of severe luminal narrowing, high-grade stenosis must be differentiated from complete occlusion [12]. Previous studies [13–17] have addressed stent artifacts and the detectability of in-stent restenosis on CT and MR angiography. Few investigations, however, have addressed the detectability of in-stent restenosis [7], especially in the important range of approximately 70–95% stenosis. In addition, some of the more recently developed and frequently used stent types have not been evaluated in this context. The purposes of this study were to use CT and MR angiography to compare frequently used stents to determine on a subjective scale the diagnosis, despite attenuation and artifact formation in the stent, of two degrees of standardized stenosis in a phantom and to objectively quantify the changes inside the stent through the use of signal intensity and attenuation profiles.
Materials and Methods
Phantom and Stents
Six stents of types of materials and structures frequently used in the management of peripheral arterial occlusive disease were placed individually in synthetic polyethylene tubes with an inner diameter of 8.5 mm. Standardized stenoses were made with solid nylon cylinders cut to 2 cm in length and bored longitudinally to a central axis of 4 or 2 mm. In this study, degree of stenosis was calculated on the basis of the area of the unobstructed lumen versus the area of the bore inside the stenosis. Compared with the unobstructed lumen inside the stents, which measured 8 mm in diameter, the degree of stenosis was 75% for the 4-mm bore and approximately 95% for the 2-mm bore. The nylon cylinders were placed inside the stents and secured with glue (Fig. 1). This test arrangement made it possible to compare stenosis without stent, stent without stenosis, and stent with stenosis. Tubes without stents but with nylon cylinders were used as controls. The tubes were plugged on the ends with silicone, and hypodermic needles were used to fill the tubes with contrast agent diluted in saline solution.
Two pieces, one with 75% and one with 95% stenosis, of each of the following stent types were used: nitinol with 1.7-mm2 mesh size and 165-μm strut thickness (Expander, Bolton Medical), nitinol with 3-mm2 mesh size and 170-μm strut thickness (Hemobahn, Prograft Medical), nitinol with 4.5-mm2 mesh size and 245-μm strut thickness (SelfX, Abbott Laboratories), nitinol with 2-mm2 mesh size and 200-μm strut thickness (SMART, Cordis), nitinol with 15-mm2 mesh size and 235-μm strut thickness (Symphony, Boston Scientific), and cobalt alloy with 1.7-mm2 mesh size and 195-μm strut thickness (Wallstent, Boston Scientific).
CT Angiography
CT angiography was performed on a 16-MDCT scanner (Somatom Sensation 16, Siemens Medical Solutions) with the following acquisition parameters: slice collimation, 16 × 0.75 mm; rotation time, 0.5 second; table feed, 14 mm/rotation; matrix size, 512 × 512. Coronal images (i.e., parallel to the table plane) were reconstructed with kernel B31 in increments of 1 and 2 mm. The tubes were filled with ioversol (Optiray 320 mg, Mallinckrodt Imaging) diluted at a ratio of 1:50 to a final concentration of 6.5 mg I/mL. To determine the concentration of contrast agent, we produced a dilution series and imaged it with CT and MRI. The concentration that most closely resembled typical attenuation or signal intensity was selected. Concentrations were adjusted such that the physiologic concentration during clinical investigation was closely approximated. Tubes were placed centrally on the gantry parallel to the longitudinal axis of the scanner with a minimum distance of 5 cm between the stents in the imaging plane to avoid additional artifacts. Previous measurements did not show field inhomogeneities even though the stents were surrounded by air and not embedded in a gelatinous solution.
MR Angiography
MR angiography (Gyroscan T10-NT Powertrak 3000, Philips Medical Systems) was performed with a 3D spoiled gradient-recalled echo sequence (TR/TE, 4.51.45; flip angle, 35°; matrix size, 160 × 256). A coronal slab with a total thickness of 96 mm was acquired, and coronal slices measuring 2 mm with 1-mm overlap were reconstructed. The tubes were filled with gadoterate dimeglumine (Dotarem, Guerbet) diluted at a ratio of 1:200 to a final concentration of 2.5 μmol/mL. To determine the concentration of contrast agent, we performed MRI on a dilution series. We compared the in vitro signal intensities with those of in vivo images and used the concentration that closely approximated in vivo values. Tubes were placed centrally in the bore parallel to the longitudinal axis of the magnet with at least 5 cm between stents to avoid radiating artifacts.
Image Analysis
DICOM images were electronically transferred to a standard PC and analyzed with freeware (Image J, National Institutes of Health). Mean pixel intensity profiles were recorded through each stenosis perpendicular to the stent axis. The dependent variable was the area under the intensity profile curve. We measured apparent degree of stenosis by comparing the percentage of pixels above a threshold signal intensity or attenuation in an artificial stenosis with vessel bore. In addition, a subjective score of 0–4 was given to each stenosis by two experienced radiologists, who were blinded to degree of in-stent stenosis and stent used. The radiologists worked in consensus to allow comparison between subjective and objective data. The subjective score was an assessment of the detectability of degree of stenosis, that is, absence of overestimation and underestimation of degree of stenosis. The true degree of stenosis was known. The measured variables are in-stent signal intensity or attenuation of unstenosed stent, average signal intensity or attenuation in the remaining channel inside the stenosed segment, and subjective score of stenosis detectability.
Results
CT Angiography
Depending on the type of stent, CT angiography showed relative in-stent attenuation for the in-stent stenoses ranging from 75% to 100% of the attenuation of the control standard (Table 1). The subjective and absolute scores of stents with 75% and 95% stenoses are summarized in Figure 2A, 2B, 2C, 2D. Stenoses in stent 3 (nitinol with 4.5-mm2 mesh size and 245-μm strut thickness, SelfX) and stent 5 (nitinol with 15-mm2 mesh size and 235-μm strut thickness, Symphony) were the most difficult to detect. These stents exhibited not only homogeneous attenuation in the lumen inside the stent but also irregular high-attenuation artifacts due to beam hardening. Depending on the artifacts induced by the mesh pattern, overestimation of in-stent stenosis was more likely in the larger-bore cases of 75% stenosis, in which the patent lumen was necessarily closer to the stent struts. In the presence of strong artifacts, the relative subjective rating was therefore lower than the objective score owing to inhomogeneous attenuation in the lumen. The overall attenuation due to the presence of the stent in unstenosed vessels is shown in Figure 3A. For both types of stenoses, strut thickness was more relevant than mesh size in terms of higher attenuation (stents 3 and 5). In a comparison of the three stents with the highest attenuation with the three stents with the lowest attenuation, there was a trend to higher mean strut thickness (220 vs 183 μm) in the stents with high attenuation on CT angiography but not high signal intensity on MR angiography. However, neither strut thickness nor mesh size proved to be a significant predictor of in-stent attenuation or subjective image quality (p > 0.13). (Figs. 2A and 2B).
| Stent Type and Angiographic Technique | Degree of Stenosis (%) | Visibility Score | Attenuation (% of Standard) |
|---|---|---|---|
| 1, Nitinol (Expander, Bolton Medical) | |||
| CT | 75 | 3 | 75 |
| CT | 95 | 4 | 91 |
| MR | 75 | 4 | 90 |
| MR | 95 | 3 | 86 |
| 2, Nitinol (Hemobahn, Prograft Medical) | |||
| CT | 75 | 3 | 77 |
| CT | 95 | 4 | 89 |
| MR | 75 | 3 | 82 |
| MR | 95 | 2 | 74 |
| 3, Nitinol (SelfX, Abbott Laboratories) | |||
| CT | 75 | 2 | 76 |
| CT | 95 | 2 | 76 |
| MR | 75 | 4 | 97 |
| MR | 95 | 4 | 92 |
| 4, Nitinol (SMART, Cordis) | |||
| CT | 75 | 3 | 93 |
| CT | 95 | 3 | 91 |
| MR | 75 | 3 | 67 |
| MR | 95 | 2 | 80 |
| 5, Nitinol (Symphony, Boston Scientific) | |||
| CT | 75 | 2 | 90 |
| CT | 95 | 2 | 83 |
| MR | 75 | 3 | 64 |
| MR | 95 | 1 | 55 |
| 6, Cobalt alloy (Wallstent, Boston Scientific) | |||
| CT | 75 | 4 | 99 |
| CT | 95 | 3 | 84 |
| MR | 75 | 1 | 33 |
| MR | 95 | 1 | 51 |
MR Angiography
On MR angiography, the relative signal intensity of the in-stent stenoses ranged from 30% to 100% of the signal intensity of the control (Table 1, Figs. 2C and 2D). The higher variation in signal intensity was mainly caused by the high signal attenuation of ferromagnetic stents (stent 6, Wallstent). Again, the presence of localized artifacts and resulting inhomogeneous luminal signal intensity caused lower subjective ratings than the subjective score would suggest. The overall signal intensity due to the presence of the stent in unstenosed vessels is shown in Figure 3B. Signal intensity correlated primarily with stent material, and there was no significant influence of strut or mesh size.
Discussion
Metallic stents are commonly used in the management of peripheral arterial stenosis [18]. In-stent restenosis, however, is one of the main reasons that long-term patency rates are low [19]. Although CT angiography and MR angiography have been successfully implemented for visualization of the peripheral arteries, visualization of the stent lumen is impeded by overall signal attenuation inside the stent [20] and localized signal inhomogeneities due to artifact formation [21, 22]. Therefore, intraarterial subtraction angiography remains the method of choice for detection of in-stent restenosis due to intimal hyperplasia or thrombus formation. In this study, we observed that the detectability of instent stenosis on CT angiography was mainly limited by overall attenuation inside the stent, whereas on MR angiography, it is dominated by both overall signal attenuation inside the stent and localized signal inhomogeneity due to artifact formation.
On CT angiography, 95% stenosis was differentiated from total occlusion inside all evaluated stent types. However, in contrast to the findings on MR angiography, low-grade stenosis tended to be underestimated owing to the presence of high-attenuation artifacts, especially in stents 3 and 5. This effect was caused in part by the relative proximity of the borders of the larger bore to the stent meshes in the case of 75% stenosis as opposed to the small and thus more centrally located bore of 95% stenosis.
The results of MR angiography were characterized by a higher variety of in-stent signal intensities across stent types. Luminal stenosis tended to be overestimated in general. In 75% stenosis, signal attenuation in the residual lumen (Fig. 3B) was mainly responsible for poorer visibility (Fig. 2C) but still allowed accurate delineation of low-grade stenosis and differentiation from fully patent vessel parts. In 95% stenosis, however, localized artifacts caused additional inhomogeneity that could have been misinterpreted as occlusion in stents 2, 4, 5, and 6.
Our experiments were conducted in a static phantom. The effect of blood flow through the various stents is expected to be negligible on CT angiography, whereas on MR angiography, flow features are known to cause changes in the appearance of stent-related artifacts due to intravoxel dephasing. However, these effects should be minor compared with susceptibility and eddy currents [23]. The stents in this experiment were scanned parallel to the z-axis. In vivo, arteries are often tortuous, and stent orientation influences the degree of artifact formation and detectability of stenosis [14]. The most important causes of in-stent restenosis are thrombus formation within the stented segment and neointimal growth through smooth-muscle cell proliferation [24–26]. Use of a central bore in our model may better approximate in vivo conditions, in which endothelial hyperplasia usually leads to concentric luminal narrowing [27].
In conclusion, CT angiography and MR angiography of in-stent stenosis suffer from different degrees of artifact formation depending on stent type and degree of stenosis. In general, CT angiography seems to be more suitable for differentiation between high-grade stenosis and occlusion, whereas MR angiography seems to have higher sensitivity in the early detection of hemodynamically relevant intermediate-grade stenosis. For stents not made of nitinol, CT angiography would generally be preferable. Nitinol stents with small meshes and thin struts tend to be better delineated on CT than on MR angiography.
Footnote
Address correspondence to M. A. Funovics ([email protected]).
References
1.
Hagspiel KD, Leung DA, Nandalur KR, et al. Contrast-enhanced MR angiography at 1.5 T after implantation of platinum stents: in vitro and in vivo comparison with conventional stent designs. AJR 2005; 184:288 –294
2.
Herborn CU, Goyen M, Quick HH, et al. Whole-body 3D MR angiography of patients with peripheral arterial occlusive disease. AJR 2004; 182:1427 –1434
3.
Heuschmid M, Krieger A, Beierlein W, et al. Assessment of peripheral arterial occlusive disease: comparison of multislice-CT angiography (MS-CTA) and intraarterial digital subtraction angiography (IA-DSA). Eur J Med Res 2003; 8:389–396
4.
Loewe C, Schillinger M, Haumer M, et al. MRA versus DSA in the assessment of occlusive disease in the aortic arch vessels: accuracy in detecting the severity, number, and length of stenoses. J Endovasc Ther 2004; 11:152 –160
5.
Zhang HL, Khilnani NM, Prince MR, et al. Diagnostic accuracy of time-resolved 2D projection MR angiography for symptomatic infrapopliteal arterial occlusive disease. AJR 2005; 184:938–947
6.
Katzberg RW, Sahler LG, Duda SW, et al. Renal handling and physiologic effects of the paramagnetic contrast medium Gd-DOTA. Invest Radiol 1990; 25:714–719
7.
Maintz D, Tombach B, Juergens KU, Weigel S, Heindel W, Fischbach R. Revealing in-stent stenoses of the iliac arteries: comparison of multidetector CT with MR angiography and digital radiographic angiography in a phantom model. AJR 2002; 179:1319 –1322
8.
Prokop M, Debatin JF. MRI contrast media: new developments and trends. CTA vs. MRA. Eur Radiol 1997; 7 [suppl 5]:299 –306
9.
Cejna M, Loewe C, Schoder M, et al. MR angiography vs CT angiography in the follow-up of nitinol stent grafts in endoluminally treated aortic aneurysms. Eur Radiol 2002; 12:2443 –2450
10.
Hong C, Chrysant GS, Woodard PK, Bae KT. Coronary artery stent patency assessed with in-stent contrast enhancement measured at multi-detector row CT angiography: initial experience. Radiology 2004; 233:286 –291
11.
Haltmayer M, Mueller T, Horvath W, Luft C, Poelz W, Haidinger D. Impact of atherosclerotic risk factors on the anatomical distribution of peripheral arterial disease. Int Angiol 2001; 20:200 –207
12.
Funovics MA, Lackner B, Cejna M, et al. Predictors of long-term results after treatment of iliac artery obliteration by transluminal angioplasty and stent deployment. Cardiovasc Intervent Radiol 2002; 25:397 –402
13.
Manke C, Nitz WR, Djavidani B, et al. MR imaging-guided stent placement in iliac arterial stenoses: a feasibility study. Radiology 2001; 219:527–534
14.
Maintz D, Kugel H, Schellhammer F, Landwehr P. In vitro evaluation of intravascular stent artifacts in three-dimensional MR angiography. Invest Radiol 2001; 36:218–224
15.
Kramer SC, Wall A, Maintz D, Bachmann R, Kugel H, Heindel W. 3.0 Tesla magnetic resonance angiography of endovascular aortic stent grafts: phantom measurements in comparison with 1.5 Tesla. Invest Radiol 2004; 39:413 –417
16.
Hamoir XL, Flohr T, Hamoir V, et al. Coronary arteries: assessment of image quality and optimal reconstruction window in retrospective ECG-gated multislice CT at 375-ms gantry rotation time. Eur Radiol 2005; 15:296 –304
17.
Baum F, Vosshenrich R, Fischer U, Castillo E, Grabbe E. Stent artifacts in 3D MR angiography: experimental studies [in German]. Rofo 2000; 172:278 –281
18.
Surowiec SM, Davies MG, Eberly SW, et al. Percutaneous angioplasty and stenting of the superficial femoral artery. J Vasc Surg 2005; 41:269 –278
19.
Faries P, Morrissey NJ, Teodorescu V, et al. Recent advances in peripheral angioplasty and stenting. Angiology 2002; 53:617 –626
20.
Sardanelli F, Zandrino F, Molinari G, Iozzelli A, Balbi M, Barsotti A. MR evaluation of coronary stents with navigator echo and breath-hold cine gradient-echo techniques. Eur Radiol 2002; 12:193 –200
21.
Bartels LW, Smits HF, Bakker CJ, Viergever MA. MR imaging of vascular stents: effects of susceptibility, flow, and radiofrequency eddy currents. J Vasc Interv Radiol 2001; 12:365–371
22.
Shetty AN, Bis KG, Kirsch M, Weintraub J, Laub G. Contrast-enhanced breath-hold three-dimensional magnetic resonance angiography in the evaluation of renal arteries: optimization of technique and pitfalls. J Magn Reson Imaging 2000; 12:912 –923
23.
Hilfiker PR, Quick HH, Debatin JF. Plain and covered stent-grafts: in vitro evaluation of characteristics at three-dimensional MR angiography. Radiology 1999; 211:693–697
24.
Cejna M, Virmani R, Jones R, et al. Biocompatibility and performance of the Wallstent and the Wallgraft, Jostent, and Hemobahn stent-grafts in a sheep model. J Vasc Interv Radiol 2002; 13:823 –830
25.
Cejna M, Virmani R, Jones R, et al. Biocompatibility and performance of the Wallstent and several covered stents in a sheep iliac artery model. J Vasc Interv Radiol 2001; 12:351–358
26.
Vorwerk D, Gunther RW, Schurmann K, Wendt G. Aortic and iliac stenoses: follow-up results of stent placement after insufficient balloon angioplasty in 118 cases. Radiology 1996; 198:45–48
27.
Yorozuya M, Suzuki H, Iso Y, et al. Comparison of the morphological changes of restenosis after the implantation of various types of stents in a swine model. Coron Artery Dis 2002; 13:305–312
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: September 25, 2006
Accepted: May 19, 2007
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.