Copyright © 2007, European Society of Cardiology
Overlapping paclitaxel-eluting stents: Long-term effects in a porcine coronary artery model
Department of Laboratory Medicine and Pathobiology, University of Toronto, Research Institute, the Hospital for Sick Children, Toronto, Ontario, Canada and Clinical Sciences Organization, Boston Scientific Corporation, Natick, Massachusetts, USA
*Corresponding author. Room 1050, 525 University Avenue, Toronto, Ontario M5G 2L3. Tel.: +1 416 813 7654x2095; fax: +1 416 813 6306. Gregory.Wilson{at}sickkids.ca
Received 14 March 2007; revised 11 June 2007; accepted 9 July 2007
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
Objective At 4-year follow-up, paclitaxel-eluting stents (PES, TAXUS®) have demonstrated clinical effectiveness in reducing restenosis without increasing death or myocardial infarction. Concerns remain with all drug-eluting stents, however, regarding potential interference with long-term healing, particularly in zones with adjacent stent overlap due to theoretical doubling in both drug release and tissue contact with coating polymer. Therefore, we evaluated long-term healing of overlapped TAXUS stents in an accepted animal model.
Methods Seventy-one non-injured swine underwent coronary artery placement of 138 overlapping stent-pairs (91 PES TAXUS Liberté 1 µg/mm2 slow release formulation, 3.0 or 3.5 mm diameter pairs and 47 control bare metal Liberté pairs) deployed at a 1.1:1 to 1.2:1 target stent-to-artery diameter ratio. Pathological analysis was performed at 30 (9 bare, 10 paclitaxel), 90 (9 bare, 10 paclitaxel), 180 (10 bare, 16 paclitaxel), 360 (10 bare, 21 paclitaxel), and 580 (9 bare, 22 paclitaxel) days.
Results At all time intervals overlapped TAXUS stents were consistently endothelialized and free of luminal thrombus or vascular dilatation. Full healing, however, was delayed compared to control, with macrophage processed para-strut fibrin and cellular debris still present, but reduced and sequestered from blood flow by an endothelialized neointima at 360 and 580 days. While neointimal thickness in TAXUS overlap zones was significantly less than control at 30 days, greater neointima formation was observed with TAXUS at 
90 days, but was stable and did not progress further from 90 to 580 days.
Conclusion In this porcine model TAXUS stents demonstrated safety and acceptable healing with prolonged time to resolution of para-strut deposits, and did not produce the sustained neointimal suppression seen clinically.
KEYWORDS Angioplasty/coronary intervention; Stents; Histopathology
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
Clinical trials of drug-eluting coronary stents (DES) have shown markedly reduced rates of clinical and angiographic restenosis compared to bare metal stents (BMS). Several randomized clinical studies with follow-up as long as four years have shown polymer-coated stents eluting sirolimus (CYPHER® Bx VELOCITY®, Cordis, Johnson & Johnson, Miami Lakes, FL) or paclitaxel (TAXUS® Express®, Boston Scientific, Natick, MA) to be effective at reducing neointimal hyperplasia with no significant increase in death or myocardial infarction [1,2]. Nevertheless, concerns remain about potential local toxicity, with delayed endothelial coverage that can contribute to a small numerical excess in very late stent thrombosis beyond 1 year compared to BMS controls. Such delayed healing may be likely at points where adjacent DES struts overlap, theoretically doubling released drug and coating polymer-tissue contact. With limited numbers of explanted DES from humans [3,4], tissue response is best assessed systematically in animal models. The few such studies published [5–11] have been constrained by short follow-up [5–7,11], use of non-clinical polymer-drug combinations [5–8], examination of non-overlapped stents [5–8,10,11], or use of peripheral (iliac) rather than coronary implant sites [7–9]. Thus, the longest reported follow-up in a porcine coronary artery model has been 180 days using single CYPHER stents [10], and the lone study of overlapped stents had only 90-day follow-up [9].
The present study comparing overlapping TAXUS stents versus BMS controls has follow-up intervals of 30, 90, 180, 360, and 580 days (including 20–21 DES examined at each of 360 and 580 days). It is thus the first study to comprehensively address long-term healing with overlapping DES in an accepted animal model [12,13].
|
2. Methods |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication 85-23, revised 1996). Stent implants were performed at Charles River Laboratories, Wisconsin Division, accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). Study procedures, including pathologic evaluation, were performed in compliance with Good Laboratory Practices (GLP) as defined in the U.S. Code of Federal Regulations, 21CFR Part 58.
2.1 Stenting protocol
Stainless steel balloon expandable stents (LibertéTM, Boston Scientific, 16-mm length, 3.0- or 3.5-mm diameter) were implanted in overlapping pairs in the left anterior descending artery (LAD), left circumflex artery (LCX), and right coronary artery (RCA) of female crossbred swine (Genetiporc, LLC, Alexandria, MN), with a 4–8 mm target overlap (double strut density) and a 24–28 mm stent-pair target length. Test (TAXUS paclitaxel-eluting stent [PES]) devices were polymer-coated with 1 µg/mm2 paclitaxel in a slow release (SR) formulation (112 µg paclitaxel per stent). Overlapping stent-pairs consisted of either 2 BMS (control) or 2 PES; each animal received 1 or 2 stent-pairs. Of 71 pigs, 47 received 1 PES and 1 BMS stent-pair, 4 received 1 PES stent-pair, and 20 received 2 PES stent-pairs. The targeted stent-to-artery diameter ratio was 1.1:1 to 1.2:1 (actual range 0.91:1 to 1.34:1) using quantitative coronary angiography, with no vessel pre-injury. Balloon pressures ranged from 8 to 18 atm. Stent type was blinded throughout the study. Follow-up was performed after 30±2 days (9 BMS, 16 PES), 90±3 days (9 BMS, 16 PES), 180±3 days (10 BMS, 16 PES), 360±7 days (10 BMS, 21 PES), and 580±25 days (9 BMS, 22 PES).
2.2 Antithrombotic regimen
Animals were pretreated with aspirin 325 mg PO and Plavix® 75 mg PO at least 3 days prior to and on the day of implantation and daily until euthanasia. Heparin (250–300 IU/kg IV plus 50,000 IU/L saline IV drip@ 100–130 mL/h) was administered at induction of anesthesia and intra-operatively for implantation and terminal procedures.
2.3 Tissue harvest and processing for histology and Scanning Electron Microscopy
At scheduled follow-up, animals were anesthetized and terminal angiography performed to document stent-pair patency. Pigs in the 360- and 580-day cohorts, however, were too large for terminal angiography. After euthanasia, hearts were harvested, labeled, and stented arteries perfusion fixed with 10% formalin overnight. Stented vessels were dissected from the myocardium, allowing adequate proximal and distal reference vessel length in addition to the stented segment, and radiographed using a high resolution Faxitron to assess for stent strut fractures. The dissected vessels along with remaining heart tissue, routine tissue samples from liver and kidneys, and any tissue in chest or abdomen which appeared abnormal, were assessed by the study pathologist (Gregory J. Wilson).
Stent-pair explants for each cohort were randomly selected for histology, scanning electron microscopy (SEM), or residual drug measurement. Table 1 lists the pairs examined at each time point. Because complete endothelialization was observed by SEM for all earlier intervals, SEM was not performed on 360- and 580-day explants.
|
Explants for histology were embedded in Spurr plastic embedding medium. All sections were stained with hematoxylin and eosin and elastic trichrome stains; in-stent sections were also stained with Carstairs' stain for fibrin. Twelve stent-pairs were histologically examined in relation to radiographically detected fractures, with additional step sections at approximately 250 µm intervals through the strut fracture zones as detected by high resolution Faxitron radiographs. The 16 formalin-fixed stent-pairs examined by SEM were bisected longitudinally into approximately equal portions, dehydrated through ethanol, and critical point dried.
Each heart was sliced transversely at approximately 1-cm intervals from base to apex and examined grossly by the pathologist for any evidence of myocardial ischemic injury or other abnormalities; suspicious areas were sectioned. In all hearts, routine transmural samples (1 or more tissue blocks) in the distribution of the RCA, LAD, and LCX downstream of the stent-pair were embedded in paraffin, sectioned at 5-µm thickness, and stained with hematoxylin and eosin. Selected unstented arteries were examined similarly for comparison. Specimens from livers, kidneys, and any other tissues submitted were grossly examined and similarly embedded, sectioned, and stained.
2.4 Histomorphological evaluation
Morphological analysis of the in-stent sections (proximal, overlap, distal) by light microscopy was performed with ordinal grading assessments of the following eight parameters (Table 2): mural thrombus, endothelialization, strut tissue coverage, para-strut leukocytes, disruption of the internal elastic lamina (IEL), disruption of the external elastic lamina (EEL), medial smooth muscle cell (SMC) loss, para-strut amorphous material (PAM). Though not formally graded, observations on in-stent sections regarding neointimal or medial hemorrhage and hemosiderin-laden macrophages, calcification, neovascularization, and prominence of eosinophils in the inflammatory infiltrates were noted. Proximal and distal non-stented reference sections were graded as 0=no abnormalities, 1=abnormality noted and described, searching particularly for any evidence of paclitaxel effects.
|
2.5 Histomorphometric evaluation
Computer-assisted morphometric analysis of the in-stent sections was performed on high-resolution (12 megapixel field) images of the stented arterial cross-sections using a combination of automated and manual techniques with calibration against a certified micrometer scale traceable to the U.S. National Bureau of Standards. Tracings provided measurements for the following (mm2): luminal area subtended by the traced lumen boundary; IEL area subtended by the inner elastic lamina boundary of the media; and EEL area subtended by the outer elastic lamina boundary of the media. Where IEL or EEL were discontinuous due to normal fenestration or disruption/loss, interpolation using existing histological features was used to provide the required continuous line tracings. From the three direct area measurements, the following derived shape-independent parameters were obtained: intimal area (IEL area–lumen area); medial area (EEL area–IEL area); percent stenosis based on IEL area: (1–[lumen area/IEL area])x100%.
Additionally, strut-to-lumen measurements from the edge of the strut farthest from the lumen to the lumen boundary were made and averaged per section to estimate intimal thickness. The strut was represented by the void it previously occupied and measurements at overlap were made from the outer layer of struts around the circumference, resting closest to the IEL, so that the distance reflected the entire neointima formed inside the IEL.
2.6 Evaluation by Scanning Electron Microscopy
The interior of the bisected halves of each stent-pair submitted for SEM were examined en face in a JEOL 820 scanning electron microscope. The entire flow surface was assessed for mural thrombus and endothelial cell coverage (flattened endothelial-like cells), examining 50 dwell points per specimen (half stent-pair) at 300x to 1000x magnification plus lower power survey views of the entire flow surface.
2.7 Histomorphologic and histomorphometric evaluation of tissue response to strut fractures
Regions from stent-pairs which showed discontinuities of one or more struts (fracture zones) were evaluated for tissue response using the morphological and morphometric parameters described in Sections 2.1 and 2.2. The focus was on fracture zone neointima formation and inflammation, evidence suggesting increased drug release, and histologic features characteristic of strut fracture, comparing between fracture zone and routine in-stent sections.
2.8 Statistical analysis
All histomorphological and histomorphometric data were statistically analyzed at Boston Scientific Corp. on a segment-by-segment basis (proximal, overlap, distal), using SAS® Version 8.02.
For morphologic data, frequency distribution of ordinal grades was reported by group (PES versus BMS) at each sectioning location and time point. For comparison of groups, morphologic values were ranked within each location and time point with one-way ANOVA performed for group differences. Similarly, differences among locations within each group and time point and differences across time points within each section were tested with P values reported. The Ryan–Einot–Gabriel–Welsh method was used to perform pair-wise comparisons on these ranks for time points within each group and location and for pair-wise comparisons of sections within each group and time point.
For morphometric parameters, a 2-sample t-test was performed to compare groups within each location and time point. One-way ANOVA was performed to test for a difference across locations within each group and time point and for comparison across time points within each group and across sectioning locations with P values reported. The Ryan–Einot–Gabriel–Welsh method was used to perform pair-wise comparisons of the time points within each group and location and for comparisons of sectioning locations within each group and time point.
P values less than 0.05 were considered significant. Two stent-pairs (PES, 90 and 580 days) were excluded from statistical analysis per the study protocol based on histological evidence of two or more struts completely outside the EEL.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
No stent-related mortality or morbidity was observed, and examination of the heart, liver, and kidneys from every animal showed no significant pathology or evidence of ischemic changes downstream of stented vessels. Reference vessel sections taken 2 mm beyond the TAXUS stents showed no evidence of paclitaxel effects (e.g., amorphous material in the intima or media).
3.1 Histomorphology and Scanning Electron Microscopy
The most informative comparisons in tissue response are presented below. Complete statistical analyses as exemplified in Tables 3A, 3B, 3C
are provided in the Online Supplement.
|
|
|
3.1.1 Luminal thrombus, strut tissue coverage, and endothelialization
Neointimal coverage of struts was complete at all time points. While the covering was often thin at 30 days, particularly in the PES group, there were no gross luminal thrombi in any stent-pair. A solitary platelet-rich microthrombus (<50 µm) was seen in the overlap zone of one 90-day PES stent-pair.
Endothelial cell coverage evaluated histologically was >90% (maximal score) in 98.8% of all sections examined. One BMS stent-pair at 30 days had <75% coverage proximally and 75%–90% coverage distally; at 580 days 75%–90% coverage was seen in one proximal and one overlap section in different PES stent-pairs. Evaluation with SEM over the entire stent-pair flow surface (en face) showed coverage approaching 100% with struts well covered in all 16 stent-pairs so examined (1 BMS and 3 PES at 30 days, 3 BMS and 3 PES at 90 days and at 180 days) (Fig. 1). To assess endothelial cell function, immunostaining for endothelial nitric oxide synthase (eNOS) was performed using the same porcine coronary artery model at 30 days in two parallel studies of PES versus BMS (non-overlapping). One study included the Liberté platform and both had the same polymer, paclitaxel loading, and release kinetics as in the current study. Complete flow surface coverage by eNOS-positive cells confirmed functional endothelium, as presented in the Online Supplement.
|
3.1.2 Inflammation (para-strut leukocytes)
Macrophages were the predominant para-strut leukocyte overall, with granulocytes and lymphocytes becoming more prevalent as inflammation increased. Grades ranged from none to severe (Table 3A), with significantly higher grades in PES versus BMS stent-pairs at 30 days (overlap; P=0.001) and 90 days (proximal [P=0.01], overlap [P=0.012], and distal [P=0.012]). At these earlier time points, no BMS stent-pair showed more than mild inflammation whereas 33.4% of PES showed moderate and 5.1% showed severe. By 180, 360, and 580 days, however, grades were not statistically different between PES and BMS and 96.2% of sections showed only mild or moderate inflammation. At 30 (P<0.001) and 180 days (P=0.034), PES groups had higher para-strut leukocyte scores in the overlap zone than in proximal and distal sections (Table 3B); there was no difference at 90, 360, or 580 days. In BMS groups, overlap scores were greater only at 180 days (P=0.003). Comparison across time points showed that the inflammatory response to PES declined significantly with longer implant time (30 through 580 days), at proximal (P=0.032) and distal (P<0.001) locations. While the decline was not significant in the overlap zones the trend (P=0.185) was towards decreasing inflammation over time (Table 3C).
The ordinal grading of para-strut leukocytes was markedly non-linear (Fig. 2); leukocyte number in severe sections was several fold greater than moderate, which was several times greater than mild. Six PES and two BMS stent-pairs were graded severe in one or more sections. Fig. 3 shows sections with combined severe inflammation and substantial numbers of eosinophils in PES and BMS groups at 360 and 580 days.
|
|
3.1.3 Internal elastic lamina and external elastic lamina disruption
Arterial wall injury is represented by the degree of disruption in continuity of the IEL and EEL. Elastica may be disrupted through direct mechanical injury from stent struts, by leukocyte elastase activity, or by matrix metalloproteinases secreted by SMC and other cell types.
Disruption of the IEL varied substantially in both PES and BMS groups. There were no significant differences at proximal, overlap, or distal sections at early (30 days) or late (360, 580 days) follow-up (Online Table 1A). Greater IEL disruption was observed at overlap in PES at 90 (P=0.034) and 180 days (P=0.018). Over time, IEL disruption (Online Table 1C) increased in BMS proximally (P=0.003), at overlap (P<0.001), and distally (P=0.011); in PES, a significant increase was observed only at overlap (P=0.034).
Disruption of the EEL was often mild or absent with no significant differences between PES and BMS (Online Table 2A). There was a trend, however, towards more instances of mild or moderate EEL disruption at overlap for PES versus BMS at 360 (P=0.09) and 580 days (P=0.063, Online Table 2A). Within the PES group at 580 days, there was significantly more EEL disruption at overlap (P=0.013) versus proximal or distal sections (Online Table 2B). Most elastica disruption/loss at overlap was well separated from struts and thus attributable to elastolysis rather than direct mechanical injury.
3.1.4 Medial smooth muscle cell loss
At all time points PES showed significantly greater medial SMC loss than BMS (Online Table 3A). This expected paclitaxel effect increased over time (Online Table 3 C), with no significant difference between overlap and non-overlap sections (Online Table 3B). No significant differences in pair-wise comparisons over time were seen at 360 and 580 days, indicating most SMC loss with PES had occurred by 90 to 180 days. Medial SMC loss with BMS also increased significantly over time and was most apparent at overlap (P<0.001, Online Table 3C). Fig. 4, which compares mild loss in BMS to the most marked observed loss (580 days, PES and BMS), illustrates the substantial medial SMC loss eventually seen in some BMS pairs.
|
3.1.5 Para-strut amorphous material
Deposition of PAM, which consisted of incompletely reabsorbed fibrin and cellular debris, was greater for PES versus BMS at all sectioning locations at 30, 90, and 180 days, reflecting persistent inhibition of cellular colonization within the para-strut region. At 360 and 580 days, this significant difference persisted in proximal and distal sections but not overlap (Online Table 4A). Among PES groups PAM was greater at overlap than in proximal or distal sections at 30 (P<0.001) and 90 days (P=0.054) but not at longer durations (Online Table 4B) and tended to decline over time (Online Table 4C). Resorption typically began in the overlap region, as may be expected given the increased inflammatory activity in this area at earlier time points (see Section 3.1.2 above); PAM remained within the neointima, covered by endothelium. Examples of PAM and its changing character over time, with fibrin and cellular debris ingested by macrophages, and the formation of small foci of microcalcification, are illustrated in Fig. 5.
|
3.2 Histomorphometry
The histomorphometric data presented below assess vascular stability and neointimal formation.
3.2.1 Vascular stability evaluation: external elastic lamina area and medial area
There were no significant increases in EEL area over time (Table 4A), suggesting the absence of vascular dilatation. The medial area showed no significant difference over time except for PES distal sections which showed a decrease (P=0.038) at 360 and 580 days (Table 4B). There were no significant differences between PES and BMS at any time point or location.
|
|
3.2.2 Neointima formation: intimal thickness and internal elastic lamina-based area percent stenosis
Intimal thickness is the most direct measure of neointima production. At 30 days, intimal thickness at overlap was significantly less with PES versus BMS and non-significant at proximal and distal locations; by 90 days there was no difference (Table 5A). At 180 days and beyond PES showed significantly thicker intima. Intimal thickness was significantly greater at overlap versus non-overlap sections for PES and BMS (all intervals but 90 days, Table 5A), likely due to the double layer of struts. Over time BMS intimal thickness declined significantly at all locations whereas with PES intimal thickness at overlap increased significantly and then reached a stable plateau over the 90–580 day time frame (Table 5B).
|
|
Area-percent stenosis, calculated from the IEL area and lumen area (Fig. 6), was significantly less (P=0.032) with PES versus BMS at overlap at 30 days. However, for the distal section at 90 days (P=0.041) and for all sections at 180 (P
0.003), 360 (P0.001), and 580 (P0.002) days PES showed significantly more stenosis than BMS, reaching a plateau over the 90–580 day time frame (not significant comparing 90, 180, 360 and 580 days) as seen with intimal thickness. Furthermore, there was no trend towards greater stenosis progressing from 90 through 580 days (Fig. 6). Overlap sections showed significantly greater stenosis versus proximal and distal sections in both groups at all intervals except 90 days.
|
3.3 Tissue response to strut fractures
Thirteen stent-pairs showed discontinuities (fractures) of one or more struts on high resolution radiographic images: 4 BMS (360 [1] and 580 [3] days) and 9 PES (30 [1], 180 [1], 360 [5], and 580 [2] days). The Faxitron imaging used in this swine study is a more sensitive detection technique compared to clinical angiographic follow-up and enabled targeted histologic evaluation of the tissue response adjacent to strut fractures. While some fracture zones showed increases in neointima compared to proximal, overlap, or distal sections from the same stent-pair, their area stenosis (
51.2%) was well within the range observed in non-fractured stent-pairs. All morphology parameters inside fracture zones were similar to those seen outside with no instances of luminal thrombus, severe inflammation, or uncovered struts. A detailed analysis of the strut fractures is presented in the Online Supplement. |
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
This is the first study to assess the long-term (up to 580 days) healing response to DES placed in an overlap configuration in swine coronary arteries. It demonstrated that overlapped TAXUS Liberté PES were consistently both well-endothelialized and free of luminal thrombus at 30 days and beyond, with no evidence of vascular dilation at all observed time intervals. Healing was delayed to 180 days in PES compared to BMS, as reflected by reduced but persistent para-strut fibrin and cellular debris that was sequestered, that is fully covered by an endothelialized neointima and gradually resolved at 360 and 580 days. Inflammatory activity and disruption of elastica (mainly elastolysis) were no longer significantly different between PES and BMS by 360 and 580 days. Neointima formation with PES was significantly greater than BMS at 90 days and beyond, but was non-progressive over the 90 to 580 day period.
The lack of sustained inhibition of neointimal hyperplasia by DES in the Yucatan mini-pig coronary model [10] for non-overlapping CYPHER stents, the rabbit iliac artery model [9] for CYPHER and TAXUS (Express® platform) overlapping stents, and the present study contrasts with the profound and sustained suppression of late loss observed in the clinic [1,2]. The moderate and stable increase in neointima with DES compared to BMS implanted under the same conditions indicates that the drug/polymer combination causes more neointimal reaction in these animal models versus the clinical situation. Carter et al. [10] reported the same failure of suppression of neointimal hyperplasia for single CYPHER stents at 90 and 180 days in porcine coronary arteries as reported here for TAXUS Liberté stents at overlap and non-overlap sections. Caution clearly is needed when extrapolating from the result of DES implantation in previously normal animal arteries to predict efficacy in diseased human arteries. This contrasts with the value of animal testing to address (though still imperfectly) safety and biocompatibility. The timing of healing is also difficult to translate from animal to human models. It has been suggested that the stages of healing are remarkably similar [14], and that 3 to 6 months in animal studies may reasonably approximate healing at 24 to 36 months in humans because the healing response in the human takes five to six times longer [6]. The present study shows temporal differences in different facets of healing, but also demonstrates full stability of the surrounding porcine coronary artery wall at approximately 1.6 years follow-up. To what extent this can predict healing in the human will require meaningful comparison with long term human explants.
In rabbit iliac arteries, Finn et al. [9] determined that delayed healing and more inflammation occurred at overlap sites of both CYPHER and TAXUS stents versus their respective BMS controls. Comparison of the two DES types suggested that TAXUS induced greater fibrin deposition, medial cell loss, heterophils/eosinophils, and late (90 days) neointimal hyperplasia at overlap, than did CYPHER, although the greater fibrin deposition resolved by 90 days. In the present study the somewhat longer persistence of para-strut fibrin was accompanied by complete neointimal coverage and endothelial cell coverage. A possible link between para-strut fibrin persistence in porcine arteries and an increase in late clinical adverse events is not supported by any differential increase in late adverse events with paclitaxel-versus sirolimus-eluting stents in large trials with patients followed up to 4 years [1].
The present study showed depletion of medial SMC in the PES groups. However, the observed stable EEL area through 580 days suggests that long-term vessel stability was not adversely impacted. Similar medial SMC depletion also occurred on occasion in overlapped BMS stent-pairs (Fig. 4). Inflammation increased early after PES placement but resolved to the point of no significant differences in either overlapped or non-overlapped sections of PES versus BMS at 180, 360, or 580 days. In fact, severe inflammation with prominent eosinophils was seen with only one PES and one BMS (Fig. 3). This is in marked contrast to the more diffuse, extreme, and persistent levels of inflammation seen in over half of CYPHER stents at 90 and 180 days in the same porcine model as reported by Schwartz and Wilson [15]. Similar inflammatory reactions have been reported in a human autopsy study of the CYPHER stent [16], although the link of such reactions to clinical events such as stent thrombosis or local aneurysm formation has not been definitively established.
There are clearly limitations in using the histopathological findings from animal models to predict clinical results with drug eluting stents, including both species differences and the absence of underlying advanced atherosclerotic disease. Nor can different drug-eluting stents be assumed to have the same time course of endothelial coverage, resorption of para-strut fibrin, inflammation, or attenuation of the underlying medial SMC population, and neointima formation. In fact, the same drug with a different polymeric carrier, dose, or release profile may show markedly different responses clinically as was seen with both polymeric and non-polymeric paclitaxel DES formulations that went on to show unsatisfactory safety profiles in clinical trials not seen with TAXUS stents. (Comparisons to some prior unsuccessful efforts at DES development are presented in the Online Supplement)
The fact remains, however, that animal models are still the best way to evaluate whether a prototype DES shows significant pro-thrombogenic or pro-inflammatory effects (safety endpoints), prior to embarking on first human use. Per this approach, the observation of persistent, severe inflammatory activity in normal animal arterial walls post stent implantation may be considered a possible, though not certain, safety concern that similar reactions will occur following human implantation, or that they will lead to clinical sequelae (such as excessive neointimal hyperplasia producing flow-limiting stenosis, pro-thrombogenicity due to inflammatory activity at the flow surface, or elastolytic-associated vascular dilatation and aneurysm formation). The observation of early inflammation-triggered neointimal proliferation, and later plateau, in the porcine model, in contrast to sustained neointimal suppression and reduction of restenosis observed clinically with TAXUS stents supports use of long-term porcine implant outcomes as an appropriately conservative safety screening model. While histopathological findings for DES in animal models should be interpreted with caution in predicting human response, they remain the best window on device safety short of clinical trials given suitable numbers, care, duration of follow-up, and evaluation of both single and overlapped stents. As such, the present study using overlapping stents shows complete endothelial coverage and an encouraging low prevalence of long term inflammation, with the TAXUS Liberté paclitaxel drug-eluting stent system.
|
Appendix A. Supplementary data |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cardiores.2007.07.004.
|
Acknowledgements |
|---|
The authors gratefully acknowledge Mohammad Eskandarian, Sue Omar, Peter Wilson, Gordana Svajger, and Cheryl Pereira in Dr. Wilson's group who assisted in pathological evaluations and manuscript preparation, including figures, together with SoJung Imm and Dr. Ruth M. Starzyk, of Boston Scientific Corporation, who performed the statistical analysis and assisted in editing the manuscript, respectively.
|
References |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Supplementary dataReferences |
|---|
- Stone G.W., Moses J.W., Ellis S.G., Schofer J., Dawkins K.D., Morice M.C., et al. Safety and efficacy of sirolimus-and paclitaxel-eluting coronary stents. N Engl J Med (2007) 356:998–1008.
[Abstract/Free Full Text] - Sousa J.E., Costa M.A., Abizaid A., Feres F., Seixas A.C., Tanajura L.F., et al. Four-year angiographic and intravascular ultrasound follow-up of patients treated with sirolimus-eluting stents. Circulation (2005) 111:2326–2329.
[Abstract/Free Full Text] - Finn A.V., Joner M., Nakazawa G., Kolodgie F., Newell J., John M.C., et al. Pathological correlates of late drug-eluting stent thrombosis. Strut coverage as a marker of endothelialization. Circulation (2007) 115:2435–2441.
[Abstract/Free Full Text] - Joner M., Finn A.V., Farb A., Mont E.K., Kolodgie F.D., Ladich E., et al. Pathology of drug-eluting stents in humans: delayed healing and late thrombotic risk. J Am Coll Cardiol (2006) 48:193–202.
[Abstract/Free Full Text] - Heldman A.W., Cheng L., Jenkins G.M., Heller P.F., Kim D.W., Ware M., et al. Paclitaxel stent coating inhibits neointimal hyperplasia at 4 weeks in a porcine model of coronary restenosis. Circulation (2001) 103:2289–2295.
[Abstract/Free Full Text] - Suzuki T., Kopia G., Gayashi S., Bailey L.R., Llanos G., Wilensky R., et al. Stent-based delivery of sirolimus reduces neointimal formation in a porcine coronary model. Circulation (2001) 104:1188–1193.
[Abstract/Free Full Text] - Klugherz B.D., Llanos G., Lieuallen W., Kopia G.A., Papandreou G., Narayan P., et al. Twenty-eight-day efficacy and pharmacokinetics of the sirolimus-eluting stent. Coron Artery Dis (2002) 13:183–188.[CrossRef][Web of Science][Medline]
- Drachman D.E., Edelman E.R., Seifert P., Groothuis A.R., Bornstein D.A., Kamath K.R., et al. Neointimal thickening after stent delivery of paclitaxel: change in composition and arrest of growth over six months. J Am Coll Cardiol (2000) 36:2325–2332.
[Abstract/Free Full Text] - Finn A.V., Kolodgie F.D., Harnek J., Guerrero L.J., Acampado E., Tefera K., et al. Differential response of delayed healing and persistent inflammation at sites of overlapping sirolimus-or paclitaxel-eluting stents. Circulation (2005) 112:270–278.
[Abstract/Free Full Text] - Carter A.J., Aggarwal M., Kopia G.A., Tio F., Tsao P.S., Kolata R., et al. Long-term effects of polymer-based, slow-release, sirolimus-eluting stents in a porcine coronary model. Cardiovasc Res (2004) 63:617–624.
[Abstract/Free Full Text] - Garcia-Touchard A., Burke S.E., Toner J.L., Cromack K., Schwartz R.S. Zotarolimus-eluting stents reduce experimental coronary artery neointimal hyperplasia after 4 weeks. Eur Heart J (2006) 27:988–993.
[Abstract/Free Full Text] - Schwartz R.S., Edelman E.R. Drug-eluting stents in preclinical studies. Recommended evaluation from a consensus group. Circulation (2002) 106:1867–1873.
[Free Full Text] - Schwartz R.S., Chronos N., Virmani R. Preclinical restenosis models and drug eluting stents. Still much to learn. JACC (2004) 44:1373–1385.
[Abstract/Free Full Text] - Virmani R., Kolodgie F.D., Farb A., Lafont A. Drug eluting stents: are human and animal studies comparable? Heart (2003) 89(2):133–138.
[Abstract/Free Full Text] - Schwartz R.S., Wilson G.J. CYPHER versus TAXUS stents: comparing the inflammatory response in porcine coronary arteries. Am J Cardiol (2006) 98(8). TCT Abstracts:36M.
- Virmani R., Guagliumi G., Farb A., Musumeci G., Grieco N., Motta T., et al. Localized hypersensitivity and late coronary thrombosis secondary to a sirolimus-eluting stent: should we be cautious? Circulation (2004) 109:701–705.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. J. Wilson, G. Nakazawa, R. S. Schwartz, B. Huibregtse, B. Poff, T. J. Herbst, D. S. Baim, and R. Virmani Comparison of Inflammatory Response After Implantation of Sirolimus- and Paclitaxel-Eluting Stents in Porcine Coronary Arteries Circulation, July 14, 2009; 120(2): 141 - 149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Drachman Drug-Eluting Stents in Animals and Patients: Where Do We Stand Today? Circulation, July 14, 2009; 120(2): 101 - 103. [Full Text] [PDF]
|
|
|
|
 |
|
 A. V. Finn, G. Nakazawa, F. D. Kolodgie, and R. Virmani Temporal Course of Neointimal Formation After Drug-Eluting Stent Placement: Is Our Understanding of Restenosis Changing? J. Am. Coll. Cardiol. Intv., April 1, 2009; 2(4): 300 - 302. [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Schwartz, E. Edelman, R. Virmani, A. Carter, J. F. Granada, G. L. Kaluza, N. A.F. Chronos, K. A. Robinson, R. Waksman, J. Weinberger, et al. Drug-Eluting Stents in Preclinical Studies: Updated Consensus Recommendations for Preclinical Evaluation Circ Cardiovasc Interv, October 1, 2008; 1(2): 143 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Shinke, J. Li, J. P. Chen, L. Pendyala, T. Goodchild, R. Jabara, S. Geva, T. Ueno, N. Chronos, K. Robinson, et al. High Incidence of Intramural Thrombus After Overlapping Paclitaxel-Eluting Stent Implantation: Angioscopic and Histopathologic Analysis in Porcine Coronary Arteries Circ Cardiovasc Interv, August 1, 2008; 1(1): 28 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Aoki, G. S. Mintz, N. J. Weissman, J. T. Mann, L. Cannon, J. Greenberg, E. Grube, A.R. Z. Masud, J. Koglin, L. Mandinov, et al. Chronic Arterial Responses to Overlapping Paclitaxel-Eluting Stents: Insights From Serial Intravascular Ultrasound Analyses in the TAXUS-V and -VI Trials J. Am. Coll. Cardiol. Intv., April 1, 2008; 1(2): 161 - 167. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||