ABSTRACT: Objective. To compare the difference of strut distribution between two clinically available drug-eluting stent platforms (Bx Velocity and Express II stents) using intravascular ultrasound (IVUS). Background. Nonuniform strut distribution (NSD) has been shown to be associated with increased intimal hyperplasia after drug-eluting stent implantation. Methods. IVUS imaging was performed on Bx Velocity (n = 6) and Express II stents (n = 6) after inflation pressures of 10, 16, and 26 atm in a bench test model. Percent NSD was defined as the length of segments with NSD (interstrut angle > 90°)/stent length. NSD was also assessed in postprocedure IVUS images in 53 clinical cases (32 Cypher, 21 Taxus) using 3-dimensional IVUS analysis. Results. Frequency of NSD segment and %NSD were lower in Bx Velocity stents than in Express II stents at the inflation pressures of 16 and 26 atm (%NSD: 16 atm, 0% vs. 13.8 ± 9.4%; p < 0.005; 26 atm, 1.1 ± 2.6% vs. 19.9 ± 6.9%; p < 0.0001). In postprocedural images from clinical cases, the frequency of NSD segment and %NSD were lower in Cypher stents than in Taxus stents (%NSD: 0.5 ± 1.6 vs. 6.8 ± 7.2; p < 0.0001). Conclusion. NSD segment was observed less in Bx Velocity stents than in Express II stents.

       Drug-eluting stents have substantially reduced the restenosis rate.^1–4 However, restenosis after drug-eluting stent (DES) implantation is resistant to repeat percutaneous coronary intervention (PCI), and remains an important clinical problem.⁵ A detailed intravascular ultrasound (IVUS) analysis has shown that nonuniform strut distribution (NSD) may be associated with increased intimal hyperplasia after DES implantation.6 However, this particular finding has not been compared among various DES types. The purpose of this study was to compare the difference in strut distribution between Bx Velocity^® (Cordis Corp., Miami, Florida) and Express II (Boston Scientific Corp., Natick, Massachusetts) stents, platforms of two U.S. Food and Drug Administration (FDA)-approved DES (Cypher™ and Taxus^®) in a bench test model and in human clinical case experience.

Materials and Methods

       Bench-top model and stent deployment. An experimental model using a silicon tube was performed on 3 x 18 mm Bx Velocity (n = 6) and 3 x 16 mm Express II stents (n = 6), as previously described.⁷ In our experiment, straight and bending (curvature 0.33 cm-1) models made of silicon tube (inner diameter of 3 mm) were used. Stents were implanted at the inflation pressure of 10 atm, and the pressure was subsequently increased to 16 and 26 atm.
       IVUS acquisition and measurements. IVUS imaging was performed following each inflation. Motorized pullback IVUS imaging was performed at 0.5 mm/second throughout the stent, with a 3.2 Fr 40 MHz IVUS imaging system (Boston Scientific). Three-dimensional IVUS volumetric analysis was performed using the EchoPlaque system (INDEC Systems, Inc., Mountain View, California).⁸ Stent volume index (SVI) was calculated as stent volume divided by stent length. The maximum angle between

Figure 1

Stent strut distribution by IVUS. White arrows point out the stent struts. The maximum angle (a) between adjacent struts was defined as interstrut angle (IA). (A) Representative image of uniform strut distribution of the Cypher™ stent. (B) Representative image of nonuniform strut distribution of the Express II stent.

adjacent struts with a protractor centered on the stent was defined as interstrut angle (IA) (Figure 1).⁷ NSD was defined as IA > 90°, and NSD segment was defined as any segment within a stent in which NSD could be continuously observed for more than 0.5 mm longitudinally. Percent NSD was defined as the length of segments with NSD divided by stent length. Protrusion was defined as prolapse of tissue between stent struts extending inside a circular arc connecting adjacent struts.
       Human clinical cases. To assess the NSD in human clinical cases, 53 patients (32 Cypher and 21 Taxus) who met the following criteria were identified from the IVUS database of the core laboratory at Stanford University (Stanford, California): (1) de novo coronary artery lesion without severe superficial calcification; (2) single stent (3.0 mm); (3) high-quality, automated pullback IVUS images at postprocedure. The frequency of NSD and %NSD in these IVUS images were assessed.
       Statistical analysis. Data are expressed as number, % or mean ± standard deviation (SD). Differences in continuous data were analyzed by the unpaired t-test. Differences in categorical data were analyzed by chi-sqaure analysis. Statistical significance was defined as p-values < 0.05. The correlation between %NSD and SVI was assessed using linear regression analysis. Statistical analysis was performed with StatView, version 5.0 (SAS Institute, Cary, North Carolina).

Results

       Bench-top model. Percent NSD was lower in Bx Velocity stents compared with Express II stents (Figure 2). In

Figure 2

Percent nonuniform strut distribution (%NSD) in Bx Velocity and Express II stents. *p < 0.005 in the Bx Velocity stent vs. the Express II stent. †p < 0.0001 in the Bx Velocity stent vs. the Express II stent.

the straight model, there was a correlation between %NSD and SVI in Express II stents (Figure 3). In Express II stents, %NSD was higher in the bending model than in the straight model at pressures of 16 and 26 atm (16 atm: 20.8 ± 6.4% vs. 6.9 ± 6.0%; p < 0.05; 26 atm: 24.8 ± 2.0% vs. 15.1 ± 6.6%; p < 0.05).
       Human clinical cases. In human cases, there was no significant difference in baseline IVUS measurements including

Table 1

stent length, SVI, maximum stent area and protrusion between the Cypher and Taxus stents (Table 1). NSD segment was observed less in Cypher stents than in Taxus stents (9.4% vs. 61.9%; p < 0.0001). Percent NSD was lower in Cypher stents compared with Taxus stents (Figure 4).

Discussion

       There were substantial differences in strut distribution between Bx Velocity and Express II stents. NSD segment was observed less in Bx Velocity stents than in Express II stents. In the bench-test model, NSD segment was observed less in Bx Velocity stents than in Express II stents at the pressures of 16 and 26 atm, confirming previous results from an in vitro study.⁷ This bench test assessed differences in strut distribution between Bx Velocity and Express II stents in not only the

Figure 3

Correlation between % nonuniform strut distribution (%NSD) and stent volume index (SVI) in Express II stents of straight model (y = 5.9x – 39.9; r = 0.86; p = 0.003).

circumferential direction, but also the longitudinal direction. Percent NSD was lower in Bx Velocity stents than in

Figure 4

Percent nonuniform strut distribution (%NSD) in Cypher™ and Taxus® stents. *p < 0.0001 in the Cypher stent vs. the Taxus stent.

Express II stents after higher-pressure expansion. Moreover, in Express II stents, %NSD was greater in the bending model than in the straight model. These data suggest that vessel curvature and stent overexpansion may result in increased NSD segment. In human cases, NSD segment was also observed less in Cypher stents than in Taxus stents, which is concordant with the experimental model.
       Stent cell design directly affects the stent strut distribution pattern. Open-cell design, as adopted by the Express II stent, may be more susceptible to NSD, especially when stents are overexpanded or implanted on a tight bend. Uniform strut distribution of the stent is theoretically important to optimize local drug delivery and to suppress intimal hyperplasia following DES implantation.^6,9 More studies to assess the NSD of DES may reveal its impact on intimal hyperplasia after DES implantation.
       Study limitations. There are several limitations in this study. First, only 3 mm diameter stents were examined in our analyses. Second, the retrospective nature of an analysis in clinical cases poses a risk for biases. Third, IVUS had a geometrical artifact due to transducer position, which might have affected strut distribution.¹⁰ Finally, the impact of NSD on intimal hyperplasia was not examined.

Conclusion

       There was a significant difference in strut distribution between Bx Velocity and Express II stents. Nonuniform strut distribution segment was observed less in Bx Velocity stents than in Express II stents.