Revealing in-stent lumen and stent artifacts in coronaries : study of in-stent lumen and stents artifacts with 64 MDCT for CT angiography Fouillé Antoine Dr Allal Mahmoudi With the precious help of Dr Sablayrolles Dr Treutenaere Dr Ohanessian Johan Cheron Radiology and interventional cardiovascular teams of CCN

Objective : Our objective was to evaluate the detectability of the in-stent lumen and artifacts of stent in coronaries using 64 slices MDCT. Materials and Methods : 19 different metallic stents (made of nitinol, cobalt, chrominium ) were implanted in plastic syringe (Nemoto SY50 of 29 mm of diameter, Japan) full of iodine contrast. The stents were dilated at different diameter. The diameter of stents is verified thank a digital photograph. CT angiography was performed on a 64-detector CT scanner (detector collimation at 64 * mm; slice thickness, 0.625 mm ; rotation speed , 0,35 sec ; gated at 67 bpm ; reconstruction, filter detail 0.625 mm every 0.3 mm). Results :

Conclusion:

CT angiography is a noninvasive alternative to digital radiographic angiography in imaging coronaries. The recent introduction of the 64-detectors CT angiography allows a better assessment of the coronaries. Metallic stents in the coronaries are known to cause artifacts that modify the intra-luminal signal. These artifacts are mainly caused by susceptibility gradients around the stent. However, we can enumerate some other kind of artifacts: • Partial volume : The partial volume appears when the structure are smaller than the voxel. The value in UH is a combination of the little structure and the environment. • Blooming effect : It appears when we scan a very dense structure.

Truncated version : Not for Medical Uses !!

Antoine Fouillé

• • •

Shadow effect : After a very dense structure, we observe a black area. Intra-Stent Enhancement : The density of the stent enhanced the density of the environment. It’s a combination of the partial volume artifact and the blooming effect. So the lumen in the stent is enhanced. Distance of influence of the stent : It is the distance after witch we can not observe an artifact of blooming of the stent.

Our objective was to evaluate the detectability of the in-stent lumen and artifacts of stent in coronaries using 64 slices CT. Materials and Methods: 1- Phantom 19 different coronary stents of different materials and designs were implanted separately in a plastic syringe (Nemoto SY50 of 29 mm of diameter, Japan) on a cotton string in the center of the syringe. The syringe was filled with a iodine contrast (Omnipaque, 350 mg I /ml, Amersham Health, Irland) mixed with isotonic solution (Bioperf 0.9%, Laboratoire Aguettant, France) in order to have a contrast medium with 13 mg I /ml. We put the stent on a cotton string on the center of the syringe.

The syringes were filled with a iodine contrast medium diluted with isotonic saline solution to 400 UH ± 10 UH, closed at the end. The stents placed in the center of the gantry. The stents were oriented parallel and then perpendicular to the z-axis of the scanner for each stent and each diameter. That allows having a maximal and a minimal value of the intra stent enhancement. 2- Stents 19 stents are used for this study : Society Boston Scientific Medtronic cordis Boston Scientific Medtronic Boston Scientific Medtronic

Model

Reference

state

Taxus Driver cypher

38001-1630 DRV30012X CRS13225

Ireland USA USA

Taxus Driver

38001-2022 DRV25012X

Ireland USA

Taxus Driver

38001-1625 DRV22512X

Ireland USA

Truncated version : Not for Medical Uses !!

Guidant Corporation Guidant Corporation

Vision

1007841-28

USA

Vision

1007829-18

USA

Medtronic

Driver

DRV27512X

USA

atrium

flyer

80152AT

USA

atrium

flyer

80153AT

USA

terumo

tsunami CSRA3530FWG Japan

atrium

flyer

80151AT

Antoine Fouillé

USA

terumo Guidant Corporation

tsunami CS-RA3030FWG Japan Vision

1007843-08

USA

cordis

velocity

VR28350

USA

cordis

crossflex 544-535

terumo

tsunami CSRA4020GWG Japan

USA

We inflate the stents in order to have the diameter of stent at 2.25, 2.5, 2.75, 2.85, 3, 3.1, 3.3, 3.5 and 4 mm if the diameters are reachable. A digital photo per stent (longitudinal) allows the measure of the exact diameter of the stent. The measure of diameter are made on the photographs with the GNU software GIMP (www.gimp.org ) and on the DICOM images with the general reformat of the software Volume Share (Advantage Workstation, GE, Milwaukee, USA). We do it for 19 stents of 6 different firms (Cordis, Boston Sc., Guidant, Medtronic, Terumo, Atrium) and 9 different diameter (2.25 mm, 2.5 mm, 2.75mm, 3mm, 3.1mm, 3.3mm, 3.5mm and 4mm of diameter). We performed 2 acquisitions per stent per diameter and we obtain 100 cases. 3- CT parameters CT was performed on a sixty four-detector scanner (VCT Lightspeed; GE, Milwaukee, USA) using a native detector collimation of 64*0.6 mm, a scan rotation time of 0.35sec, a pitch: 0.24, a tube current of 650 mA, and a tube voltage of 120 kV. The scan field of view of 32 cm was used (small). The acquisition is gated at 67 bpm to simulate a cardiac acquisition. A display field of view of 96 mm with a 512*512 pixels matrix was chosen, resulting in an in-plane resolution of approximately 0.19*0.19 mm. Volume data sets with an effective slice thickness of 0.625 mm and a 50% reconstruction overlap were obtained using the scanner’s standard adaptive axial image reconstruction algorithm and a detail kernel. The mode segment of reconstruction was used. In addition, multiplanar reformations in the longitudinal axis of the stents were created (thickness, 0.625 mm). 4- Evaluation of intra-stent lumina and artifacts Axial CT images and longitudinal CT reformations were assessed with the general reformat of the software Volume Share (Advantage Workstation, GE, Milwaukee, USA) . CT angiograms were displayed with a window width of 800 UH and a center of 200 UH. Four measures were taken for statistics: 1- Distance of influence of the stent (in mm) : A profil going through the center of the ROI and at least one node of the stent was plot. We measured the distances between the maximum of intensity and the average intensity of the solution without the stent. For measures per stent and per diameter. The value provided is an average of all the measure for a stent. Truncated version : Not for Medical Uses !!

Antoine Fouillé

2- Enhancement in the center of the stent (in UH): We plot a profil in the center of the stent and take the average value (standard deviation less than 15 UH). The enhancement is a difference between the average value in the center of the stent and the average value of the solution without the stent.

3- Maximum density of the stent (in UH) : a 3D ROI was performed on the stent, the maximum value was taken. An exception was made for Terumos stents, the 3D ROI was made without the marker. 4- Diameter of the stent (in mm) : We placed a circular ROI in order to have the center of the stent. An average of three measures going through the center of the ROI and at least one node of the stent was calculated.

Truncated version : Not for Medical Uses !!

Antoine Fouillé

One binary decision was taken • We check if the intra-stent lumina is visible or not. The diameter was asses with a digital camera (Nikon D70, Japan) and an 18-70 objective (Nikor AF-S DX, Nikon, Japan) and had a resolution of 6.1 effective Mega pixels. The RAW data were exported to jpeg format without losses. The error margin is ± 100 µm, that means less than 5 %.

5- Measures and statistic Comparison for one in vitro diameter of different stents ( 2.25 mm at 3 mm and 3 mm at nomimal diameter for example) • Test of χ² for the visibility of the intrastent lumen. • Type of artifact per stent • Influence of artifact (distance in mm) : (1) • Density in UH of the stent : (3) • Enhancement of the intrastent lumen (2)

Results The lumen of all the stents is visible with an appropriate window. All the stents provide artifacts. The lumen of all the stent tsunami we scan is too artifacted at the level of the marker for a diagnosis. Comparison between all the measure of diameter: • Correlation between diameter measure with scan and announce diameter for all the stents: 0.9

Truncated version : Not for Medical Uses !!

Antoine Fouillé

• •

Correlation between diameter measure with scan and diameter measure with the photograph for all the stents: 0.9 Correlation between diameter measure with the photograph and announce diameter for all the stents: 0.9

• Coronary 64-slice CTA is a recently introduced method for noninvasive imaging of the coronary arteries with a further improvement of spatial and temporal resolution compared with the previous 16-slice CT. As stents of different material and design exhibit different lumen visibility in CT, our intention was to create a catalogue of the CT appearance of most currently available coronary stents with a diameter under 3.5 mm. Our results confirm the high variability of stent lumen visibility depending on stent type that was reported previously for 4-slice and 16-slice CT [3].

mark

model

Guidant Corporation Vision Boston Sc Taxus Medtronic Driver terumo tsunami Medtronic Driver cordis crossflex Guidant Corporation Vision cordis velocity Medtronic Driver terumo tsunami Medtronic Driver terumo tsunami atrium flyer atrium flyer Boston Sc Taxus Boston Sc Taxus atrium flyer cordis cypher Guidant Corporation Vision classement des stent

Cypher

Velocity

zone of influence (in standard deviation mm)

nominal diameter 2.72 2.5 2.25 4 3 3.5 3.5 3.5 2.75 3 2.5 3.5 2.5 3.5 3 2.25 3 2.25 2.5 Crossflex

1.07 1.07 1.1 1.13 1.14 1.14 1.15 1.16 1.17 1.17 1.17 1.19 1.21 1.22 1.23 1.26 1.26 1.3 1.34 Vision

reference

0.04 1007841-28 0.07 38001-1625 0 DRV22512X 0.07 CS-RA40JP20GWG 0.1 DRV30012X 0.12 544-535 0.09 1007843-08 0.09 VR28350 0.04 DRV27512X 0.05 CS-RA3030FWG 0.07 DRV25012X 0.15 CS-RA3530FWG 0.07 80151AT 0.06 80153AT 0.05 38001-1630 0.12 38001-2022 0.19 80152AT 0.07 CRS13225 0.18 1007829-18 Flyer

Driver

Taxus

The lumen in the center of the stent is not constant. The structure of the stent is indeed not constant and so influences the density in the center of the lumen in different ways. That standard deviation is between 5 and 15 UH in the center of the stent. Conclusion The angiosurgery are used to the stent the more dense. However, for follow up of the intra stent re-stenosis and patency, we recommande to use the less dense stent.

Truncated version : Not for Medical Uses !!

Antoine Fouillé

Tsunami

The 2D analysis, and especially the cross section is a reference tool for analyze of stents. The lumen in the stent and outside are different. The analyze should be done separately.

References 1- David Maintz, Bernd Tombach, Kai-Uwe Juergens, Stefanie Weigel, Walter Heindel, Roman Fischbach (2002) Revealing In-Stent Stenoses of the Iliac Arteries: Comparison of Multidetector CT with MR Angiography and Digital Radiographic Angiography in a Phantom Model. Key words: stent, CT, multidetector, iliac arteries, MDCT, MR, Digital radiography, wax.

2- Gillard Martine Gilard M, Cornily JC, Rioufol G, Finet G, Pennec PY, Mansourati J, et al (2005) Noninvasive assessment of left main coronary stent patency with 16slice computed tomography. Am J Cardiol 95(1):110–112 Key words: stent, CT, 16-slices, MDCT, multidetector, patency, CTangiography.

3- David Maintz, Harald Seifarth, Rainer Raupach, Thomas Flohr, Michael Rink, Torsten Sommer, Murat Özgün, Walter Heindel, Roman Fischbach (2005) 64-slice multidetector coronary CT angiography: in vitro evaluation of 68 different stents. Eur Radiol (2006) 16: 818–826 Key words: CT angiography, stent, 64-slices, artifacts,cornanary

4- Michael S. Lee, MD, Varinder Singh, MD, Thomas J. Nero, MD, James R. Wilentz, MD (2002) Cutting Balloon Angioplasty study. J Invasive Cardiol 14(9):552-556, 2002 Key words: Cutting Balloon Angioplasty, In-Stent Restenosis,

5- Maintz D, Botnar RM, Fischbach R, Heindel W, Manning WJ, Stuber M (2002) Coronary magnetic resonance angiography for assessment of the stent lumen: a phantom study. J Cardiovasc Magn Reson 4(3):359–367 6- Ligabue G, Rossi R, Ratti C, Favali M, Modena MG, Romagnoli R (2004) Noninvasive evaluation of coronary artery stents patency after PTCA: role of Multislice Computed Tomography. Radiol Med (Torino) 108(1–2):128–137 7- Mahnken AH, Buecker A, Wildberger JE, Ruebben A, Stanzel S, Vogt F, et al (2004) Coronary artery stents in multislice computed tomography: in vitro artifact evaluation. Invest Radiol 39 (1):27–33

Truncated version : Not for Medical Uses !!

Antoine Fouillé