Testing procedure for lumen quantification of in-stent restenosis using magnetic resonance imaging

Objetives: This study evaluated a test procedure for quantification of in-stent restenosis under magnet resonance imaging (MRI).

Methods: 3 carotid stents (30–34×8mm) have been investigated consisting of different materials and stent designs (Strecker Tantalum (open cell), Boston Scientific; Nitinol Precise RX and 316L Palmaz Genesis, both (closed cell), Cordis). An appropriate material (non-disclosed due to confidentiality) for simulating the artificial in-stent restenosis has been pre-selected by comparison of different materials T1 parameter with vascular tissue of a pig's aorta at 1 and 1.5 Tesla. The stents were placed in silicone tubes. Artificial stenosis of 50%, 75% and 90% of the diameter were built into the stents. Reference tubes without stents were used. The tubes were positioned in a static phantom. All were filled with 0.9% NaCl solution. Stents were placed with axes parallel to the static magnetic field into the head coil of the 1 Tesla MR system (Intera, Philips Medical Systems). We ran gradientecho, T1- & T2-weighted spinecho (SE) and 3D balanced Fast Field Echo (b3DFFE) pulse sequences. Signal intensities in- and outside the stents were measured in the MR image using DICOM tools. Artifacts were rated qualitatively. In a second test with same sequences, the NaCl solution inside the tubes was doped with a clinical dose (0.2 mmol/kg) of contrast medium (Gadovist, Schering, Berlin, Germany).

Results: According to each stent type different results appeared (Fig 1). The lumen was covered by artifacts in the 316L stainless steel stent. The Nitinol stent reduced the signal of the rest lumen, thus no evaluation was possible. In spite of signal reduction the artificial restenosis were visible in the T2-SE only. 50% artificial stenosis resulted in 43%; 75% stenosis owned a degree of 69%; 90% stenosis was depicted with 81% of the diameter. The Tantalum stent produced the lowest artifacts, thus measurements of the different stenosis were possible. Using the b3DFFE, the Tantalum stent was depicted with actual value at 50% and 75% degrees of stenosis. Merely with 90% there was a divergence of 2.5% resulting in 87.5% depicted stenosis.

Conclusion: This test procedure allows a quantitative determination of in-stent restenosis by analysis with MRI. It is obvious, that the stent material has great influences on the visibility of an in-stent restenosis. With the choice of a low-artifact stent material and a design with low RF shielding, a quantification of an in-stent restenosis is possible.