Meta-Analysis of Angiographic Versus Intravascular Ultrasound Parameters of Drug-Eluting Stent Efficacy (from TAXUS IV, V, and VI)

Meta-Analysis of Angiographic Versus Intravascular Ultrasound Parameters of Drug-Eluting Stent Efficacy (from TAXUS IV, V, and VI) Esteban Escolar, MDa, Gary S. Mintz, MDb, Jeffrey Popma, MDc, Aleksandra Michalek, MDa, Sang Wook Kim, MD, PhDa, Lazar Mandinov, MD, PhDd, Joerg Koglin, MD, PhDd, Gregg Stone, MDb, Stephen G. Ellis, MDe, Eberhard Grube, MDf, Keith D. Dawkins, MDg, and Neil J. Weissman, MDa,* Both quantitative coronary angiography (QCA) and intravascular ultrasound (IVUS) are currently used to assess in-stent restenosis. This study aimed to use standardized imaging and clinical follow-up to compare QCA parameters with several IVUS parameters to evaluate their strengths and weaknesses for detecting in-stent restenosis in a drug-eluting stent population. A subset of patients from the TAXUS IV, V, and VI studies was evaluated. The subset, which included 216 TAXUS-treated patients and 191 bare-metal stent–treated patients, had complete IVUS and QCA performed at baseline and follow-up. As expected, all QCA and IVUS parameters were consistent with less intimal hyperplasia in TAXUS patients than controls. The overall incidence of QCA binary restenosis was 14.0%, which was 9.3% in TAXUS-treated patients and 19% in bare-metal stent–treated patients (p ⴝ 0.0008). Regression analysis showed that QCA late lumen loss and percentage of diameter stenosis correlated only moderately with the various IVUS measures of neointimal hyperplasia for the combined group of patients (TAXUS ⴙ bare-metal stent), as well as for the TAXUS-treated and bare-metal stent–treated patients separately. However, in general, correlations within the control (bare-metal stent) group tended to be stronger than within the TAXUS group. The strongest correlation was between QCA percentage of diameter stenosis and IVUS percentage of intimal hyperplasia in the overall group and the control group. The strongest IVUS predictor of QCA binary restenosis at 9 months was maximum percentage of intimal hyperplasia, with an overall C ⴝ 0.91 and p <0.001. In conclusion, the QCA and IVUS parameters used to evaluate drug-eluting stent efficacy showed a moderate correlation with IVUS percentage of intimal hyperplasia, reliably predicting QCA binary in-stent restenosis. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;100:621– 626)

Drug-eluting stents (DESs) have dramatically decreased, but not totally eliminated, in-stent restenosis. The primary quantitative end point for assessing the effectiveness of DESs typically has been quantitative coronary angiography (QCA); specifically, measures of late loss, follow-up diameter stenosis, and binary restenosis.1 However, these angiographic measures are surrogates for intimal hyperplasia (the biologic process of in-stent restenosis). Conversely, intravascular ultrasound (IVUS) can visualize and measure ina

Cardiovascular Research Institute/Medstar Research Institute, Washington Hospital Center, Washington, DC; bCardiovascular Research Foundation, New York, New York; cDepartment of Cardiology, Brigham and Women’s Hospital, Boston; dBoston Scientific Corp, Natick, Massachusetts; eThe Cleveland Clinic Foundation, Cleveland, Ohio; fHeart Center, Siegburg, Germany; and gSouthampton University Hospital, Southampton, United Kingdom. Manuscript received October 31, 2006; revised manuscript received March 4, 2007, and accepted March 20, 2007. *Corresponding author: Tel.: 202-877-0223; fax: 202-877-0206. E-mail address: [email protected] (N.J. Weissman). Dr. Mandinov and Dr. Koglin are full-time employees of Boston Scientific Corporation, Marlborough, Massachusetts. 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.03.076

timal hyperplasia and therefore directly quantify the in-stent restenosis process. Using angiographic and IVUS data from the Treatment of denovo coronary disease using a single pAclitaXel eluting stent studies (TAXUS) randomized controlled trials, we compared multiple angiographic and IVUS measures of restenosis to better understand the relative strengths and weaknesses of these parameters. Methods Patient population and protocol: The TAXUS trials that used the polymer-based paclitaxel-eluting TAXUS Express stent (Boston Scientific, Natick, Massachusetts) in de novo lesions (TAXUS IV, V de novo, and VI) were described previously.2 All 3 studies were double-blind randomized controlled trials. The institutional review board at each participating center approved each of the 3 studies, and consecutive eligible patients signed informed written consent. For the purpose of this report, we analyzed only patients with complete angiographic and volumetric IVUS data at both time of implantation and 9-month follow-up. The current analysis therefore includes a subset of 216 TAXUS-treated patients (68 from TAXUS IV, 104 from www.AJConline.org

622

The American Journal of Cardiology (www.AJConline.org)

Figure 1. Illustration of IVUS measurement of maximal intimal hyperplasia (IH) thickness. Once the lumen and the stent are traced, the software detects the centroid of the stent and constructs 180 lines through this centroid. The computer calculates the maximal distance between the lumen border and stent, the maximum IH thickness. Average distance between the lumen border and stent represents the mean IH thickness. CSA ⫽ cross-sectional area.

Figure 2. Illustration of IVUS measurement of percentage of neointimal free length of stent (%NFLS). The number of cross-sectional image slices (1 mm apart) with no detectable neointimal hyperplasia using IVUS is divided by total length of the stent. Abbreviations: EEM ⫽ external elastic membrane; IH ⫽ intimal hyperplasia.

TAXUS V, and 44 from TAXUS VI) and 191 bare-metal stent–treated control patients (68 from TAXUS IV, 84 from TAXUS V, and 39 from TAXUS VI).

ameter stenosis at follow-up), and late lumen loss (calculated as minimal lumen diameter after the procedure ⫺ follow-up minimal lumen diameter) were calculated.3

Angiography analysis: Independent QCA analysis was performed at the angiographic core laboratory by a technician who was unaware of patients’ clinical outcomes. A single angiographic core laboratory was used for all patients in this report. The in-stent analysis encompassed only the segment covered by the stent, whereas the in-lesion analysis also included 5-mm segments proximal and distal to the stent edge. Minimal lumen and reference diameters were measured after the procedure and at follow-up. Percentage of diameter stenosis, restenosis rate (defined as ⬎50% di-

IVUS protocol and analysis: After intracoronary administration of nitroglycerin 0.1 to 0.2 mg, IVUS imaging was performed using a motorized transducer pullback system (0.5 mm/s) and commercial catheters (Atlantis SR, Boston Scientific, Maple Grove, Minnesota, or Eagle Eye or Avanar F/X, Volcano Corp., Rancho Cordova, California). Images were stored on s-VHS tapes or compact disc in Digital Imaging and Communications in Medicine (DICOM) format. Quantitative volumetric IVUS analysis was performed at an independent core laboratory by a technician who was

Coronary Artery Disease/Imaging Assessment of DES Efficacy

623

Table 1 Selected baseline patient characteristics Variable

Baseline clinical characteristic Age (yrs) Women Diabetes mellitus Requiring medication Requiring insulin Hypertension requiring medication Hyperlipidemia requiring medication Current smoking Previous myocardial infarction Clinical presentation Unstable angina pectoris Stable angina pectoris Lesion characteristics Left anterior descending Lesion type A B1 B2 C Total stent length (mm)

TAXUS IV, V, VI (n ⫽ 216)

Control IV, V, VI (n ⫽ 191)

p Value

TAXUS IV (TAXUS ⫹ control) (n ⫽ 36)

TAXUS V (TAXUS ⫹ control) (n ⫽ 188)

TAXUS VI (TAXUS ⫹ control) (n ⫽ 83)

p* Value

62.7 ⫾ 10.3 55 (25.5%) 62 (28.7%) 44 (20.4%) 18 (8.3%) 146 (67.4%)

62.9 ⫾ 10.1 64 (33.5%) 45 (23.6%) 25 (13.1%) 20 (10.5%) 127 (66.5%)

0.83 0.08 0.26 0.06 0.49 0.91

62.7 ⫾ 10.2 35 (25.7%) 31 (22.8%) 22 (15.9%) 8 (5.9%) 102 (75%)

62.1 ⫾ 10.5 62 (33%) 61 (32.4%) 38 (20%) 26 (13.8%) 120 (63.6%)

64.6 ⫾ 9.6 22 (26.5%) 15 (18.1%) 15 (18.2%) 4 (4.8%) 51 (61.4%)

0.18 0.31 0.02 0.59 0.01 0.04

155 (71.8%)

126 (65.8%)

0.23

93 (68.2%)

130 (69%)

58 (69.9%)

0.96

51 (23.7%) 69 (32.1%)

37 (19.6%) 53 (27.9%)

0.33 0.33

29 (21.3%) 37 (27.2%)

39 (20.7%) 58 (30.6%)

21 (25%) 28 (33.7%)

0.71 0.59

65 (30.1%) 122 (56.7%)

62 (32.5%) 103 (54%)

0.66 0.54

42 (30.9%) 75 (55%)

69 (36.7%) 96 (51%)

16 (19.3%) 56 (68%)

0.015 0.03

97 (44.7%)

81 (42.6%)

0.66

67 (49%)

71 (37.8%)

42 (50.4%)

0.01

17 (7.9%) 65 (30.2%) 74 (34.4%) 59 (27.4%) 27.37 ⫾ 12.02

12 (6.3%) 36 (18.8%) 83 (43.5%) 60 (31.4%) 26.83 ⫾ 11.8

0.56 0.008 0.06 0.38 0.59

15 (11.1%) 55 (40.7%) 49 (36.3%) 16 (11.9%) 18.53 ⫾ 5.03

13 (6.9%) 29 (15.5%) 76 (40.4%) 70 (37.2%) 28.15 ⫾ 12.74

1 (1.2%) 17 (20.5%) 32 (38.6%) 33 (39.8%) 27.18 ⫾ 3.94

0.0148 0.0001 0.76 0.0001 0.0001

Values expressed as mean ⫾ SD or number (percent). * p Value for baseline comparison across all studies.

unaware of patients’ clinical outcomes. A single IVUS core laboratory was used for all patients in this report. With the use of computerized planimetry (TapeMeasure, Indec Systems, Mountain View, California), stent and lumen edges were manually traced. Stent, lumen, and intimal hyperplasia (stent ⫺ lumen) areas were measured every 1 mm within the stented segment. Volumes were calculated using Simpson’s rule. Percentage of intimal hyperplasia was defined as intimal hyperplasia volume divided by stent volume. Maximum percentage of intimal hyperplasia was defined as the largest percentage of intimal hyperplasia (intimal hyperplasia area divided by stent area) at any point within the length of the stent. IVUS binary restenosis was defined as maximum percentage of intimal hyperplasia ⬎50%. Minimum lumen area was defined as the smallest lumen area within the length of the stent. Angiographic diameter stenosis is a comparison of 2 linear dimensions with a goal of obtaining a surrogate measure of intimal hyperplasia thickness within the stent. Intimal hyperplasia thickness on IVUS therefore was also measured to determine its relative value. TapeMeasure software automatically places 180 lines passing through the centroid of the mass of the lumen contour and the entire stent circumference. Intimal hyperplasia thickness was measured as the distance between the lumen and stent along each of these lines (Figure 1). Maximal, mean, and minimal intimal hyperplasia thicknesses were calculated.4,5 Percentage of neointimal free length of the stent was measured as the

length of each stent that was free of IVUS-detectable neointima divided by the total length of the stent (Figure 2). Statistical analysis: Continuous variables are presented as mean ⫾ 1 SD. Factorial design analyses using 2-way analysis of variance testing treatment and study interaction for each QCA and IVUS parameter were performed showing no significant interaction p values for all variables. This justifies the pooling and comparison of TAXUS and baremetal stents using simple t test. Adjusted R-square by correlation based on partial correlation for the overall population was performed. For correlation assessment in the TAXUS and bare-metal stent group only, partial correlation was sought by adjusting for study. Statistical significance was set at p ⬍0.05. Results There were some minor differences in baseline clinical characteristics among the 3 TAXUS trials (Table 1). Mean IVUS stent areas were equal in the TAXUS arms among the 3 studies (7.80 ⫾ 1.87 in TAXUS IV, 7.70 ⫾ 2.41 in TAXUS V, and 7.90 ⫾ 1.7 mm2 in TAXUS VI; p ⫽ 0.22). Percentages of intimal hyperplasia were also similar in the TAXUS arms among the 3 studies (11.12 ⫾ 12.32% in TAXUS IV, 13.77 ⫾ 12.75% in TAXUS V, and 11.03 ⫾ 11.48% in TAXUS VI; p ⫽ 0.31). Percentage of neointimal free length of the stent was shorter in the TAXUS arm of TAXUS V compared with TAXUS IV and VI (40.31 ⫾

624

The American Journal of Cardiology (www.AJConline.org)

Table 2 Quantitative coronary angiography (QCA) and intravascular ultrasound (IVUS) analysis at 9-month follow-up

QCA (in-stent) Minimum lumen diameter (mm) Late lumen loss (mm) Diameter stenosis (%) IVUS (in-stent) Mean stent CSA (mm2) Minimum lumen CSA (mm2) Mean lumen CSA (mm2) IH volume (%) Mean IH CSA (mm2) Maximum IH thickness (mm) Mean IH thickness (mm) NFLS (%)

TAXUS IV, V, VI (n ⫽ 216)

Control IV, V, VI (n ⫽ 191)

p Value*

2.24 ⫾ 0.6 0.41 ⫾ 0.56 17.93 ⫾ 18.75

1.84 ⫾ 0.64 0.86 ⫾ 0.51 34.91 ⫾ 18.12

0.0001 0.0001 0.0001

7.51 ⫾ 2.27 5.34 ⫾ 2.28 6.58 ⫾ 2.29 12.49 ⫾ 12.42 0.93 ⫾ 0.99 0.47 ⫾ 0.35 0.11 ⫾ 0.12 48.8 ⫾ 35.98

8.01 ⫾ 2.44 4.58 ⫾ 2.34 5.7 ⫾ 2.25 29.54 ⫾ 14 2.32 ⫾ 1.35 0.77 ⫾ 0.31 0.27 ⫾ 0.15 13.4 ⫾ 22.14

0.03 0.0010 0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001

* p Value for treatment comparison between TAXUS and control based on pooled data. CSA ⫽ cross-sectional area; IH ⫽ intimal hyperplasia; NFLS ⫽ neointimal free length of stent. Table 3 Correlation between quantitative coronary angiography (QCA) late lumen loss (LLL) and intravascular ultrasound (IVUS) measures of intimal hyperplasia

Table 4 Correlation between quantitative coronary angiography (QCA) percentage of diameter stenosis (DS) and intravascular ultrasound (IVUS) measures of intimal hyperplasia

IVUS parameter

IVUS Parameter

Intimal hyperplasia (%) Overall TAXUS Control Neointimal free lenght of stent (%) Overall TAXUS Control Mean intimal hyperplasia thickness Overall TAXUS Control Maximum intimal hyperplasia thickness Overall TAXUS Control

R Correlating QCA LLL to the IVUS Parameter

p Value

0.59 0.57 0.63

⬍0.0001 ⬍0.0001 ⬍0.0001

⫺0.36 ⫺0.38 ⫺0.32

⬍0.0001 ⬍0.0001 ⬍0.0001

0.55 0.57 0.56

⬍0.0001 ⬍0.0001 ⬍0.0001

0.52 0.47 0.61

⬍0.0001 ⬍0.0001 ⬍0.0001

33.65% vs 57.24 ⫾ 37% in TAXUS IV and 59.57 ⫾ 34.11% in TAXUS VI; p ⫽ 0.0019). Maximum intimal hyperplasia thickness was greater in the TAXUS arm of TAXUS V compared with TAXUS IV (0.52 ⫾ 0.31 vs 0.40 ⫾ 0.33 mm; p ⫽ 0.01), but was similar between TAXUS V and VI. QCA and IVUS parameters were similar among the 3 bare-metal stent control groups. Baseline characteristics of the TAXUS- and bare-metal stent–treated groups were similar (Table 1). QCA reference vessel diameter was smaller for the TAXUS group than the bare-metal stent group (2.75 ⫾ 0.45 vs 2.87 ⫾ 0.54 mm; p ⫽ 0.01), although baseline minimal lumen diameters were similar. At follow-up, QCA minimal lumen diameter, late lumen loss, and percentage of diameter stenosis in TAXUS-treated patients were all consistent with less intimal hyperplasia (Table 2). Also at follow-up, IVUS mean and minimum lumen areas were larger and all mea-

Intimal hyperplasia (%) Overall TAXUS Control Neointimal free lenght of stent (%) Overall TAXUS Control Mean intimal hyperplasia thickness Overall TAXUS Control Maximum intimal hyperplasia thickness Overall TAXUS Control

R Correlating QCA p Value %DS to IVUS Parameter 0.52 0.49 0.56

⬍0.0001 ⬍0.0001 ⬍0.0001

⫺0.23 ⫺0.26 ⫺0.24

⬍0.0001 0.0002 0.002

0.45 0.44 0.46

⬍0.0001 ⬍0.0001 ⬍0.0001

0.43 0.39 0.49

⬍0.0001 ⬍0.0001 ⬍0.0001

sures of intimal hyperplasia were smaller in TAXUS-treated patients (Table 2). Almost half the length of the stent was free of neointima in the TAXUS-treated group compared with 13.4% neointima-free length in the bare-metal stent controls. QCA versus IVUS correlations: Regression analysis showed that QCA late lumen loss (Table 3) and percentage of diameter stenosis (Table 4) correlated only moderately with the various IVUS measures of neointimal hyperplasia for the combined group of patients (TAXUS ⫹ bare-metal stent), as well as for the TAXUS-treated and bare-metal stent–treated patients separately. However, correlations within the control (bare-metal stent) group tended in general to be stronger than within the TAXUS group. The strongest correlation was seen between QCA percentage of diameter stenosis and IVUS percentage of intimal hyperplasia in the overall group and the control group.

Coronary Artery Disease/Imaging Assessment of DES Efficacy

625

Table 5 Overall (TAXUS ⫹ bare-metal stent) intravascular ultrasound predictors for 9-month quantitative coronary angiography binary restenosis Parameter

Maximum IH thickness % IH Minimum Lumen CSA % Neointimal free length of stent Maximum % IH

Control

TAXUS

Overall

C Statistic

p Value

In-Stent (%DS ⬎50)

C Statistic

p Value

In-Stent (%DS ⬎50)

C Statistic

p Value

In-Stent (%DS ⬎50)

0.829 0.860 0.794 0.636

0.7392 0.6498 0.0001 0.5999

22.73% 24.1% 22.16% 24.1%

0.843 0.849 0.689 0.763

0.5613 0.6905 0.0002 0.5136

8.87% 9.23% 8.49% 9.23%

0.856 0.876 0.777 0.740

0.5206 0.6552 ⬍0.0001 0.3467

15.3% 16.1% 14.9% 16.1%

0.929

⬍0.0001

22.16%

0.883

0.0279

8.49%

0.917

⬍0.0001

14.9%

DS ⫽ diameter stenosis; other abbreviations as in Table 2.

IVUS predictors of QCA binary restenosis: The overall incidence of QCA binary restenosis was 14.0%, which was 9.3% in TAXUS-treated patients and 19.0% in baremetal stent–treated patients (p ⫽ 0.0008). The strongest predictor of QCA binary restenosis at 9 months was IVUS maximum percentage of intimal hyperplasia, with an overall C ⫽ 0.91 and p ⬍0.001 (Table 5). QCA predictors of IVUS binary restenosis: The overall incidence of IVUS binary restenosis (maximum percentage of intimal hyperplasia ⬎50%) was 30%, which was 15% in TAXUS-treated patients and 46% in bare-metal stent–treated patients (p ⬍0.0001). None of the QCA parameters tested had the power to predict IVUS binary restenosis. Discussion Although they assess the same biologic process, the present study showed only a moderate relation between QCA and IVUS parameters of stent restenosis. The current metaanalysis combined angiographic and IVUS data from the TAXUS trials. Use of a standardized angiographic and IVUS acquisition protocol and single core angiographic and IVUS laboratories across the trials provided the consistency necessary for meta-analysis. QCA late lumen loss is the end point used most often in randomized clinical DES trials.6 – 8 However, it represents the difference between 2 minimal lumen diameter measurements at 2 different times; the axial location of the minimal lumen diameter at each time point is variable. Thus, it is not a true measure of maximum intimal hyperplasia thickness within the stent. Accordingly, this analysis showed that late lumen loss correlated with maximum intimal hyperplasia thickness with an r value of only 0.52 overall, 0.47 for TAXUS stents and 0.61 for bare-metal stents. A previously reported QCA analysis from TAXUS IV showed a good relation between target lesion revascularization and late lumen loss, but only if late lumen loss was ⬎0.6 mm.9 Stents have been compared on the basis of their angiographic restenosis rates, easily differentiating DESs from control bare-metal stents.1,7,8 However, a comparison of different DESs may be more problematic because there is often little difference in QCA restenosis rates. IVUS percentage of intimal hyperplasia as a continuous variable was higher than QCA percentage of diameter stenosis. Similarly,

the rate of IVUS binary restenosis was also higher than angiographic binary restenosis. Using this measure, fewer patients would be required to detect a difference between 2 DESs. As an example, using this IVUS definition of binary restenosis to compare a DES with a bare-metal stent (based on differences in percentage of intimal hyperplasia ⬎50% between TAXUS and bare-metal stent), it would be necessary to have only 42 patients per group to detect a difference with a statistical power of 80% and confidence interval of 95%. In the same-case scenario, but using angiographic binary restenosis as a surrogate end point, 129 patients per group would be necessary. If the intention is to compare 2 different DESs, using the IVUS definition of binary restenosis (and assuming an absolute difference between the groups of 10% as an extrapolation from the differences between Cypher [Cordis Corporation, Miami, Florida] and TAXUS stents in the Randomized Comparison Between CYPHER and TAXUS Stents [REALITY] and Sirolimus-Eluting Stent Compared with Paclitaxel-Eluting Stent for Coronary Revascularization [SYRTAX] studies10), it would be necessary to have 189 patients per group to show a statistical difference with a power of 80% and confidence interval of 95% with IVUS. If an angiographic definition of binary restenosis is used to compare 2 different DESs with these same assumptions, 640 patients per group would be necessary. In recent clinical trials, most DES restenoses were focal.11,12 Focal in-stent restenosis is the result of focal accumulation of neointimal hyperplasia. Thus, maximum percentage of intimal hyperplasia should have a high degree of accuracy to detect focal in-stent restenosis. Conversely, classic IVUS measures of in-stent restenosis, percentage of intimal hyperplasia volume (often called the net volume obstruction), minimize the impact of the more common focal intimal hyperplasia accumulation. This was also shown by the analysis of percentage of neointimal free length of the stent. In the TAXUS group, almost half the stent was free of IVUS-detectable intimal hyperplasia. The present study was a retrospective subanalysis of the TAXUS IV, V, and VI studies. There were some differences in study populations among the protocols; however, given the objective of this substudy, those differences did not affect our analysis. IVUS was also conducted in only a subset of angiographic patients and thus there was the potential for selection bias.

626

The American Journal of Cardiology (www.AJConline.org)

Acknowledgment: We thank Peter Lam, PhD, Boston Scientific Corp., Maple Grove, Minnesota, for assistance with statistical analyses. 1. Mauri L, Orav EJ, Kuntz RE. Late loss in lumen diameter and binary restenosis for drug-eluting stent comparison. Circulation 2005;111: 3435–3442. 2. Stone GW, Ellis SG, Cox DA, Hermiller J, O’Shaughnessy C, Mann JT, Turco M, Caputo R, Bergin P, Greenberg J, Popma JJ, Russell ME. A polymer-based paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med 2004;350:221–231. 3. van der Zwet PM, Reiber JH. A new approach for the quantification of complex lesion morphology: the gradient field transform: basic principles and validation results. J Am Coll Cardiol 1994;24:216 –224. 4. Hoffmann R, Mintz GS, Pichard AD, Kent KM, Satler LF, Leon MB. Intimal hyperplasia thickness at follow-up is independent of stent size, a serial intravascular ultrasound study. Am J Cardiol 1998;82:1168 – 1172. 5. Hoffmann R, Mintz GS, Haager PK, Bozoglu T, Grube E, Gross M, Beythien C, Mudra H, vom Dahl J, Hanrath P. Relation of stent design and stent surface material to subsequent in-stent intimal hyperplasia in coronary arteries determined by intravascular ultrasound. Am J Cardiol 2002;89:1360 –1364. 6. Mintz GS, Hong MK, Raizner AE, Lee CW, Kim JJ, Escolar E, Fearnot NE, Park SW, Park SJ, Weissman NJ. Intravascular ultrasound assessment of neointima distribution and the length of stent that was free of intravascular ultrasound-detectable intimal hyperplasia in paclitaxel-eluting stents. Am J Cardiol 2005;95:107–109.

7. Lansky AJ, Costa RA, Mintz GS, Tsuchiya Y, Midei M, Cox DA, O’Shaughnessy C, Applegate RA, Cannon LA, Mooney M, et al, for the DELIVER Clinical Trial Investigators. Non-polymer-based paclitaxel-coated coronary stents for the treatment of patients with de novo coronary lesions: angiographic follow-up of the DELIVER clinical trial. Circulation 2004;109:1948 –1954. 8. Moses JW, Leonv MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, Jaeger JL, Kuntz RE, for the SIRIUS Investigators. Sirolimuseluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–1323. 9. Ellis SG, Popma JJ, Lasala JM, Koglin JJ, Cox DA, Hermiller J, O’Shaughnessy C, Mann JT, Turco M, Caputo R, et al. Relationship between angiographic late loss and target lesion revascularization after coronary stent implantation analysis from the TAXUS-IV trial. J Am Coll Cardiol 2005;45:1193–1200. 10. Silber S. Cypher versus Taxus: are there differences? J Interv Cardiol 2005;18:441– 446 11. Colombo A, Orlic D, Stankovic G, Corvaja N, Spanos V, Montorfano M, Liistro F, Carlino M, Airoldi F, Chieffo A, Di Mario C. Preliminary observations regarding angiographic pattern of restenosis after rapamycin-eluting stent implantation. Circulation 2003; 107:2178 –2180. 12. Weissman NJ, Koglin J, Cox DA, Hermiller J, O’Shaughnessy C, Mann JT, Turco M, Caputo R, Bergin P, Greenberg J, et al. Polymerbased paclitaxel-eluting stents reduce in-stent neointimal tissue proliferation: a serial volumetric intravascular ultrasound analysis from the TAXUS-IV trial. J Am Coll Cardiol. 2005;45:1201–5.