In-vitro evaluation of coronary stents and 64-detector-row computed tomography using a newly developed model of coronary artery stenosis.

BACKGROUND: Stent implantation is the predominant therapy for non-surgical myocardial revascularization in patients with coronary artery disease. However, despite substantial advances in multidetector computed tomography (MDCT) coronary imaging, a reliable detection of coronary in-stent restenosis is currently not possible. PURPOSE: To examine the ability of 64-detector-row CT to detect and to grade in-stent stenosis in coronary stents using a newly developed ex-vivo vessel phantom with a realistic CT density pattern, artificial stenosis, and a thorax phantom. MATERIAL AND METHODS: Four different stents (Liberté and Lunar ROX, Boston Scientific; Driver, Medtronic; Multi-Link Vision, Guidant) were examined. The stents were placed on a polymer tube with a diameter of 2.5, 3.0, 3.5, or 4.0 mm. Different degrees of stenosis (0%, 30%, 50%, 70-80%) were created inside the tube. For quantitative analysis, attenuation values were measured in the non-stenotic vessel outside the stent, in the non-stenotic vessel inside the stent, and in the stenotic area inside the stent. The grade of stenosis was visually assessed by two observers. RESULTS: All stents led to artificial reduction of attenuation, the least degree of which was found in the Liberté stent (11.3+/-10.2 HU) and the Multi-Link Vision stent (17.6+/-17.9 HU; P = 0.25). Overall, the non-stenotic vessel was correctly diagnosed in 55.5%, the low-grade stenosis in 58.3%, the intermediate stenosis in 63.8%, and the high-grade stenosis in 80.5%. In the 3.0-, 3.5-, and 4.0-mm vessels, in none of the cases was a non-stenotic or low-grade stenotic vessel misdiagnosed as intermediate or high-grade stenosis. The average deviation from the real grade of stenosis was 0.40 for the Liberté stent, 0.46 for the Lunar ROX stent, 0.45 for the Driver stent, and 0.58 for the Multi-Link Vision stent. CONCLUSION: Our ex-vivo data show that non-stenotic stents and low-grade in-stent stenosis can be reliably differentiated from intermediate and high-grade in-stent stenosis in vessels with a diameter of 3 to 4 mm. With regard to artifacts and the grading of stenoses, the Liberté stent was best suited for CT coronary angiography.

In vitro evaluation of coronary stents and in-stent stenosis using a dynamic cardiac phantom and a 64-detector row CT scanner.

INTRODUCTION: The aim of the study was to examine the ability of a 64-slice MDCT to detect in-stent stenoses in an ex vivo model of coronary stents. METHODS: Five different stents (Liberté, Boston Scientific; Driver, Medtronic; Multi-Link Vision, Guidant; Taxus Express, Boston Scientific; Cypher, Cordis) were examined using a dynamic cardiac phantom. The stents were pulled over a vessel model that consists of a polymer tube with diameters of 3.0, 3.5, and 4.0 mm and four different degrees of stenoses (0%; 30%; 50%; 70-80%). This model was moved with a rate of 60 bpm to mimic cardiac motion. To assess the degree of artificial signal reduction (artificial reduction of attenuation (ARA)) by the different stents, attenuation values were measured in the vessel outside the stent, and in the non-stenotic vessel inside the stent. Furthermore the grade of stenosis was assessed by two clinical observers. RESULTS: Highest ARA was found for the Cypher Stent (35 HU), whereas the Liberté Stent presented the lowest ARA (16 HU). Depending on the stent and the vessel diameter, up to 87.5% of the stenoses were correctly diagnosed. In the 3.0 and 3.5 mm vessels, a nonstenotic or low-grade stenotic vessel was diagnosed as intermediate or high-grade stenosis in 22.5%, whereas in the 4.0 mm vessels, this kind of overestimation did not occur. A 50% stenosis was diagnosed as a 30% stenosis in 30%. On the other hand, high-grade stenoses were underestimated in only 10%. On a four-point scale, the average deviation from the real grade of stenosis was 0.21 for the Liberté stent, 0.54 for the Taxus Express stent, 0.29 for Driver stent, 0.62 for the Multi-Link Vision stent, and 0.37 for the Cypher stent. CONCLUSIONS: In a dynamic cardiac phantom model, high grade stenoses in vessels with a diameter of 4 mm could be reliably detected irrespective of the stent type used in this study. Vice versa, high grade stenoses (> or = 50%) could only be ruled out with certainty in vessels with a diameter of 4 mm. In smaller vessels, the ability to correctly diagnose high-grade stenoses was dependent on the type of stent and the imaging artifacts associated with it.