Coronary Interventions

Transcription

1 Coronary Interventions Long-Term Safety of an Everolimus-Eluting Bioresorbable Vascular Scaffold and the Cobalt-Chromium XIENCE V Stent in a Porcine Coronary Artery Model Fumiyuki Otsuka, MD, PhD; Erica Pacheco, MS; Laura E.L. Perkins, DVM, PhD; Jennifer P. Lane, BS; Qing Wang, MD, PhD; Marika Kamberi, PhD; Michael Frie, BS; Jin Wang, PhD; Kenichi Sakakura, MD; Kazuyuki Yahagi, MD; Elena Ladich, MD; Richard J. Rapoza, PhD; Frank D. Kolodgie, PhD; Renu Virmani, MD Background The Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb) has shown promising clinical results; however, only limited preclinical data have been published. We sought to investigate detailed pathological responses to the Absorb versus XIENCE V (XV) in a porcine coronary model with duration of implant extending from 1 to 4 months. Methods and Results A total of 335 devices (63 Absorb and 7 XV) were implanted in or 3 main coronary arteries of 136 nonatherosclerotic swine and examined by light microscopy, scanning electron microscopy, pharmacokinetics, and gel permeation chromatography analyses at various time points. Vascular responses to Absorb and XV were largely comparable at all time points, with struts being sequestered within the neointima. Inflammation was mild to moderate (with absence of inflammation at 1 month) for both devices, although the scores were greater in Absorb at 6 to 36 months. Percent area stenosis was significantly greater in Absorb than XV at all time points except at 3 months. The extent of fibrin deposition was similar between Absorb and XV, which peaked at 1 month and decreased rapidly thereafter. Histomorphometry showed expansile remodeling of Absorb-implanted arteries starting after 1 months, and lumen area was significantly greater in Absorb than XV at 36 and 4 months. These changes correlated with dismantling of Absorb seen after 1 months. Gel permeation chromatography analysis confirmed that degradation of Absorb was complete by 36 months. Conclusions Absorb demonstrates comparable long-term safety to XV in porcine coronary arteries with mild to moderate inflammation. Although Absorb was associated with greater percent stenosis relative to XV, expansile remodeling was observed after 1 months in Absorb with significantly greater lumen area at 36 months. Resorption is considered complete at 36 months. (Circ Cardiovasc Interv. 14;7:33-34.) The current metallic drug-eluting stents (DESs) have been shown to exhibit several limitations including a risk for hypersensitivity reaction, late stent thrombosis, and the development of neoatherosclerosis as well as an impaired vasomotion and inability to assess lumen dimensions by noninvasive imaging. 1 3 The Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb; Abbott Vascular, Santa Clara, CA) is designed to provide transient arterial support while eluting an antiproliferative drug. The first-in-man ABSORB Cohort A trial demonstrated a low rate of major adverse cardiac events (3.3%) at 1 year with restored vasomotion and late lumen enlargement at years in patients with a single de novo native coronary artery lesion, 4,5 although the potential benefit of these outcomes over metallic DESs needs to be proven in randomized clinical trials. The ABSORB Cohort B trial also showed safety and efficacy with major adverse cardiac event rate of Key Words: coronary disease pathology stents 9.% at years. 6 Although these clinical studies on Absorb have been conducted in a relatively small number of patients with simple coronary lesions, large-scale randomized clinical trials on the second-generation everolimus-eluting cobaltchromium XIENCE V (XV) stents (Abbott Vascular) in more complex lesions have shown clinical safety with similar but lower major adverse cardiac event rates: 7.1% at years in the SPIRIT (Clinical Evaluation of the XIENCE V Everolimus Eluting Coronary Stent System) IV trial. 7 Safety of XV is further supported by human autopsy results, with a low rate of stent thrombosis (4%) at late and very late time points. 8 The current preclinical study sought to investigate detailed pathological changes after the implant of Absorb as compared with XV in a porcine coronary model extending from 1 to 4 months (total devices evaluated: 61 Absorb). Quantitative coronary angiography assessment was performed along with Received October 17, 13; accepted May 7, 14. From the CVPath Institute, Inc, Gaithersburg, MD (F.O., E.P., K.S., K.Y., E.L., F.D.K., R.V.); Abbott Vascular, Santa Clara, CA (L.E.L.P., J.P.L., Q.W., M.K., J.W., R.J.P.); and American Preclinical Services, Minneapolis, MN (M.F.). The Data Supplement is available at Correspondence to Renu Virmani, MD, CVPath Institute, Inc, 19 Firstfield Rd, Gaithersburg, MD rvirmani@cvpath.org 14 American Heart Association, Inc. Circ Cardiovasc Interv is available at DOI: /CIRCINTERVENTIONS

2 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold 331 WHAT IS KNOWN Clinical studies have demonstrated the safety and efficacy of the Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb) in a relatively small number of patients with simple coronary lesions; however, only limited preclinical data have been reported. Large-scale randomized clinical trials involving more complex coronary lesions, along with human pathological studies, have shown the safety of the second-generation everolimus-eluting cobalt-chromium XIENCE V (XV) stents with a low rate of late and very late stent thrombosis. WHAT THE STUDY ADDS Absorb demonstrated comparable safety to XV in a normal porcine coronary artery model 4 months after implantation with low-grade inflammation and fibrin deposition. Inflammation scores were greater in Absorb compared with XV at 6 to 36 months; however, inflammation scores for Absorb decreased with time. Histomorphometric analyses showed late lumen and vessel enlargement in Absorb-implanted arteries starting after 1 months. Histomorphologic changes of Absorb dismantling were observed after 1 months, and resorption sites of Absorb were poorly discernible to essentially absent at 4 months. Gel permeation chromatography analysis confirmed that degradation of Absorb was complete by 36 months. pharmacokinetics and gel permeation chromatography (GPC) analyses at various time points to correlate the histological observations with the scaffold s performance in a preclinical model. Methods Study Devices The Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb) is the same construct used in Cohort B of the ABSORB clinical trial. Absorb is a balloon-expandable, bioresorbable scaffold that consists of a poly(l-lactide) backbone coated with a poly(d, l-lactide) coating in a 1:1 ratio with everolimus. XV is a balloon-expandable cobalt-chromium stent that served as the metallic DES control for this study. Absorb and XV both have the same everolimus dose density (1 μg/cm ). 9 Notably, postimplantation structural components of Absorb are referred to as struts from time of implantation to 3 months, whereas the term resorption site is applied to denote the location of pre-existing structural components of Absorb at 36 months. Resorption site is defined as hydration and replacement of the scaffold strut by infiltrating proteinaceous material and connective tissue. Experimental Model The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee, and the study was conducted in compliance with the Food and Drug Administration Good Laboratory Practice regulations. Fifty nonatherosclerotic healthy Yorkshire crossbred farm swine ( 3 months) and 86 Yucatan mini swine (6 4 months) underwent device implantation with a targeted balloon-to-artery ratio of 1. to 1.1 (for details, see Methods in the Data Supplement). Each animal received a single Absorb or XV implant (3. 18 mm for 1, 3, and 6 months and 3. 1 mm for 1 to 4 months) in or 3 main coronary arteries. Absorb and XV were implanted according to a predetermined stratified matrix, and animals were randomly assigned to the matrix. Consequently, a total of 335 devices (63 Absorb and 7 XV) were implanted in 136 animals. One Yorkshire crossbred farm swine with Absorb and 1 XV died 1 hour postimplantation from unknown cause and was excluded from the analysis; however, the implants were intact without defect or thrombus. Seventy-one animals were euthanized and examined at 1 (n=19 devices), 3 (n=1), 6 (n=1), 1 (n=1), 18 (n=19), 4 (n=19), 3 (n=), 36 (n=), and 4 months (n=1) by light microscopy or scanning electron microscopy (SEM; 11 Absorb and 71 XV; Table 1). In addition to light microscopy and SEM analysis, pharmacokinetic studies were performed in 4 animals (53 Absorb) and GPC analysis in 4 animals (98 Absorb) at various time points (Tables 1 and ). Quantitative coronary angiography assessment was also performed on all animals designated for light microscopy and SEM. The details of the methods for quantitative coronary angiography, SEM, pharmacokinetics, and GPC analyses are described in Methods in the Data Supplement. Histological Preparation Hearts were excised and perfusion fixed with 1% neutral buffered formalin at 8 to 1 mm Hg. The implanted coronary arteries were dissected from the heart, radiographed, dehydrated in graded series of ethanol solutions, and embedded in methyl methacrylate plastic. After polymerization, - to 3-mm sections were sawed, and 3 segments from the proximal, mid, and distal portions of each device were sectioned. Histological sections were cut at 6 μm and stained with hematoxylin and eosin and Movat pentachrome for light microscopic assessment. Histological Analysis Histological sections were measured with computer assist software (IPLab, Scanalytics, Rockville, MD) as previously described. 1 Measurements included the cross-sectional areas for external elastic lamina, internal elastic lamina (IEL), and lumen. The percent stenosis was calculated using the following formula: (1 [lumen area/iel area]) 1. The results of external elastic lamina, IEL, and lumen area measurements are also being reported in a separate manuscript. 11 Neointimal thickness was measured and defined as the distance from the inner surface of each strut/resorption site to the luminal border. Because Absorb resorption sites become difficult to identify over time, estimates of neointimal thickness were used at later time points (3 4 months). Giant cell reaction was expressed as a percentage of the total number of struts/resorption sites in each section. Vessel injury score was scored according to the method by Schwartz et al. 1 Neointimal inflammation ( 4) and fibrin deposition ( 3) were semiquantitatively scored for each section as previously described. 8,13 Statistical Analysis Results for continuous variables with normal distribution were expressed as mean±sd. Normality of distribution was tested by the Shapiro Wilk test. Variables with non-normal distribution were expressed as median and 5th to 75th percentiles. Integrated linear generalized estimating equation modeling with an assumed Gaussian distribution, an identity link function, and a compound symmetry structure for working correlation matrix was used for the comparisons of continuous variables between Absorb and XV to account for the longitudinal structure of device usage within the same animals. Dependent variables with non-normal distribution were logarithmically transformed before the analyses. All analyses were performed with JMP 5 (SAS Institute Inc, Cary, NC) and SAS version 9. (SAS Institute Inc). The statistical tests were tailed, and a value of P<.5 was considered to be of statistical interest.

3 33 Circ Cardiovasc Interv June 14 Table 1. Summary of Studied Animals and Implanted Materials for Light Microscopy, SEM, and GPC 1 mo 3 mo 6 mo 1 mo 18 mo 4 mo 3 mo 36 mo 4 mo Total Light microscopy Pigs 6* Implants Absorb 1* XIENCE V 6* SEM Pigs Implants Absorb 8 XIENCE V GPC Pigs Implants Absorb Quantitative coronary angiography assessment was performed on all animals for light microscopy and SEM. Absorb indicates Absorb bioresorbable vascular scaffold; GPC, gel permeation chromatography; and SEM, scanning electron microscopy. *Twelve Absorb and 7 XIENCE V were implanted in 7 pigs for light microscopic assessment at 1 month; however, 1 animal with Absorb and 1 XIENCE V died 1 hour postimplantation from unknown cause and was excluded from the analysis. One animal with 1 Absorb and 1 XIENCE V for light microscopic assessment at 4 months was euthanized at 4 months because of health issues; however, the vascular responses observed in this animal was comparable with that observed in other animals examined at 4 months and therefore the animal was included into the 4 months group. Results Angiographic Analysis Coronary angiography demonstrated mild luminal narrowing at 1 month and moderate luminal narrowing at 3 months for both Absorb and XV; however, the degree of luminal narrowing became much less thereafter, and at 4 months, smooth lumen contours were observed in both devices. Balloonto-artery ratio in each group remained similar for all time points, although XV showed greater balloon-to-artery ratio as compared with Absorb at each time point (Table 3). Both Absorb and XV tended to show larger follow-up mean luminal diameter at later time points as compared with earlier time points with the exception of 36 months at which time both Absorb and XV were of similar size to <18 months. Late lumen loss and percent diameter stenosis were greater for Absorb as compared with XV at 1 to 6 months, followed by comparable late loss and percent diameter stenosis between the groups at 1 to 4 months. Absorb exhibited lower late loss and smaller percent diameter stenosis as compared with XV from 3 to 4 months, where the difference was statistically significant at 36 months. Late loss and percent diameter stenosis were maximal at 3 months for both Absorb and XV but decreased thereafter for both implants 1 months, and Table. Summary of Studied Animals and Implanted Materials for Pharmacokinetics Pharmacokinetics 3 h 1 d 3 d 7 d 14 d 1 mo mo 3 mo Total Pigs Implants Absorb Absorb indicates Absorb bioresorbable vascular scaffold. the decrease was more remarkable for Absorb than XV after 1 months (Table 3). Histological Analysis Light microscopic assessment revealed that vascular responses to Absorb were largely comparable with those to XV at all time points (Figures 1 and ) with patent stents/scaffold and near (1 month) to complete (>1 month) sequestration of struts within the neointima. Neither Absorb nor XV showed evidence of luminal thrombosis of either the main or the side branch of coronary arteries. Morphometric analysis showed expansile vessel remodeling in Absorb-implanted arteries starting at 1 months, whereas XV showed no changes in external elastic lamina or IEL areas (Figure 3, Table 4). Increased vessel area in Absorb was observed after 1 months, was most dramatic at 18 months, and followed at a slower rate of enlargement after 18 months. In contrast, XV showed minimal change in vessel size throughout the period of observation. The lumen area paralleled the changes observed in IEL. The percent area stenosis was significantly greater in Absorb at all time points except at 3 months as compared with XV and peaked at 3 to 6 months for both Absorb and XV. Neointimal thickness was also significantly greater in Absorb than XV at all time points except at 3 months. However, the lumen area enlarged after 1 months, which was only seen in Absorb, and the difference in lumen area between the devices reached statistical significance at 36 and 4 months. At 36 and 4 months, Absorb showed greater percent area stenosis (6.4±6.1% and 4.5±4.9%, respectively) as compared with XV (19.4±8.8% and 16.±1.6%, respectively); however, the effective lumen was larger in Absorb as compared with XV.

4 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold 333 Table 3. Quantitative Coronary Angiography 1 mo 3 mo 6 mo 1 mo 18 mo 4 mo 3 mo 36 mo 4 mo P Value* Injury scores were generally low and comparable within the groups 3 months, although the score was greater for Absorb as compared with XV despite lower balloon-to-artery ratio in the former, and the difference was statistically significant at 1 months (Figure 3, Table 4). Inflammation was overall mild to moderate for both devices, with absence of inflammation at 1 month in Absorb and XV. However, granulomas were observed in 3 of 1 in Absorb (1 at 3 months and at 1 months) and in 4 of 67 XV (3 at 3 months and 1 at 1 months). Although inflammation scores were significantly greater for Absorb than XV from 6 to 36 months except for 4 months, the scores progressively declined after 18 months in Absorb (P<.1) for the comparison across time (18 4 months). Both devices exhibited absent or minimal inflammation at 4 months. After exclusion of the stents/scaffold with granulomas, the inflammatory infiltrate was minimal to mild and consisted mostly of macrophages and a few lymphocytes (Figures and 3). Giant cells were the most frequently observed at 1 month in both devices; however, they were uncommon after 3 months, and there was no significant difference between the devices (Figures and 3). The extent of fibrin deposition was similar in both groups; peak fibrin deposition was seen at 1 month with rapid decrease P Value for Interaction Balloon-to-artery ratio Absorb 1.3±.6 1.6±.3 1.7±.4 1.8±.6 1.5±.5 1.5±.7 1.5±.3 1.5±.4 1.7± XIENCE V 1.11±.6 1.1±.5 1.1± ± ± ±. 1.11± ± ±.4.79 P value Preimplant mean luminal diameter, mm Absorb.81±.19 3.±.1.91±..7±.15.71±.17.73±.18.85±.18.74±.11.71± XIENCE V.77±.19.85±.1.86±.1.68±.16.63±.7.67±.9.87±..79±.18.69±.15.6 P value Postimplant mean luminal diameter, mm Absorb.89±. 3.11±.1.98±.16.78±.19.84±.18.81±..95±.15.81±.1.8± XIENCE V.9±. 3.9±.9 3.9± ±.15.91±.14.9± ±.17.97±..88±.1.98 P value Follow-up mean luminal diameter, mm Absorb.31±.19.4±.18.9±.19.49±.18.8±.3.84± ±.37.65±.9 3.±.4 <.1.1 XIENCE V.67±.3.34±.4.71±.6.77±.3.78± ±.6 3.8±.3.58±.3 3.5±.9.4 P value Late lumen loss, mm Absorb.58±.1.87±.17.69±.18.9±.17.±.36.3±.33.6±.4.15±.31.4±.35 <.1 <.1 XIENCE V.8±.11.74±.41.38±.15.4±.18.13±.17.11±.5.1±..39±.6.1±.31 <.1 P value Percent diameter stenosis, % Absorb 19.85± ± ± ±5.8.34± ± ± ± ±1.6 <.1 <.1 XIENCE V 9.54± ± ± ± ± ± ± ± ±1.5 <.1 P value Absorb indicates Absorb bioresorbable vascular scaffold. *P value for the comparison across all time points. P value for interaction with time and device. P value for Absorb versus XIENCE V at each time point. at 3 months and becoming absent or minimal at 6 months (Figures and 3). Minor areas of calcification were identified around struts/resorption sites of Absorb at all time points (Figure ), although these are interpreted to relate to preexisting poly(d, l-lactide) coating and are likely an inconsequential feature. There were time-dependent aspects appreciated with struts/resorption sites of Absorb, both of which are a reflection of the scaffold s transience: increasing observation of discontinuities and eosinophilic staining of the resorption sites by hematoxylin and eosin and blue staining by Movat stain. Discontinuities, defined as sites in which struts/resorption sites deviated from the conventional box shape with integration of arterial derived tissue, were increasingly observed with time, especially at time points after 1 months. In addition, beginning at 18 months, resorption sites underwent color changes that included alcian blue positive staining at 18 and 4 months with Movat pentachrome and increasing eosinophilia with hematoxylin and eosin thereafter (Figures 1 and, respectively). In concert with both of these chronological changes, there was increasing integration of arterial tissue into resorption sites such that by 4 months sites were often completely replaced by collagen and proteoglycans with sparse

5 334 Circ Cardiovasc Interv June 14 A 1 month 3 months 6 months 1 months Absorb 18 months 4 months 3 months 36 months 4 months B 1 month 3 months 6 months 1 months XIENCE V 18 months 4 months 3 months 36 months 4 months Figure 1. Representative histological images of Absorb bioresorbable vascular scaffold (Absorb; A) and XIENCE V (B) in porcine coronary arteries from 1 to 4 months (Movat pentachrome). Color changes are appreciated at resorption sites beginning at 18 months in that resorption sites are increasingly alcian blue positive from 18 to 4 months. Thereafter, resorption sites undergo further changes, and there is progressive dismantling with corresponding integration of proteoglycans and collagen with few smooth muscle cells. The resorption sites are almost completely replaced by connective tissue at 4 months. spindle cells (eg, smooth muscle cells [SMCs]) and rare remnants of calcification. Although all arteries implanted with Absorb and XV were patent without evidence of thrombus formation, there were arteries (1 each at 1 and 3 months) implanted with Absorb in which overlaid struts, struts incompletely apposed to the IEL, and strut discontinuities were observed (Figure 4). In the artery evaluated at 1 month, these observations were present in the mid to distal sections of the artery, whereas similar observations were located proximally in the artery evaluated at 3

6 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold months 1 month 6 months 1 months 18 months 4 months 3 months 36 months 4 months Figure. Histological images showing minimal to mild inflammation, fibrin deposition, and calcification around Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb) struts/resorption sites (hematoxylin and eosin stains). Mild inflammatory infiltrate was observed from 1 to 36 months (black arrows at 1 and 18 months) and consisted mostly of macrophages and a few lymphocytes. At 4 months, inflammation was not readily observed. Giant cells were occasionally observed at 1 month (black arrowheads); however, they were uncommon after 3 months. Peak fibrin deposition was noted at 1 month (white arrowheads) with rapid decrease at 3 months and becoming absent or minimal thereafter. Minor areas of calcification (white arrows) were observed around struts/resorption sites after 3 months. months. Both Absorbs had been implanted in the left anterior descending artery and had received a second inflation proximally to ensure apposition. Both arteries had near (1 month) to complete (3 months) neointimal coverage of all struts, including those struts that deviated from their expected location. The maximum percent stenosis as assessed histologically in these arteries at 1 and 3 months were 39% (distal) and 65% (proximal), respectively, which related to the increased neointimal proliferation that occurred in relation to the overlaid struts. In neither artery was excess neointimal or adventitial inflammation nor injury to the artery wall observed; however, fibrin was observed at these sites. Additional evaluation of the myocardium of these animals showed no evidence of thromboembolization in relation to the site of implant. Scanning Electron Microscopy At 1 month, both Absorb and XV showed complete strut coverage with mild neointimal growth such that the underlying outline of the struts could be clearly identified. The luminal surface appeared 1% endothelialized with cells arranged in cobblestone morphology and showing generally tight and well-formed cell-to-cell contacts (Figure 5). Rapid endothelial coverage observed at 1 month was maintained to the final SEM follow-up at 1 months for both Absorb and XV (Figure 5); however, only devices for Absorb and 1 for XV were evaluated at each time point. Pharmacokinetics The mean cumulative percentage of everolimus released during the first 8 days was 79%, with 35% being released during the first 4 hours. Everolimus release peaked at 9 days (96%) after implantation of Absorb (Figure 6A). The maximum everolimus concentration in scaffolded arterial segments occurred at 3 hours after implantation with a mean of 16. ng/mg (Figure 6B). For the effectiveness of inhibition of human SMC proliferation, everolimus concentrations in the target vessels for the first 8 days postimplant should be maintained at levels minimally greater than the IC 5 of SMC. 14 The results showed that everolimus arterial concentrations ranged from.3 to 4.6 ng/mg from 1 to 8 days, which were greater than the IC 5 of SMC. The levels decreased to.6 ng/mg by 9 days postimplantation. Peak levels of everolimus in the adjoining nonscaffolded segments were.7% of that in scaffolded segments in the proximal region and 1.7% in the distal region. The peak concentration (C max ) of everolimus in blood was 3.75±.96 ng/ml at 15 minutes after deployment of three mm Absorbs in 1 animal. For systemic safety, it is desirable that the drug concentrations in blood are at levels <3 ng/ml which is the minimally effective concentration for immunosuppression in organ transplant patients. 15 The blood levels of everolimus quickly declined to below the limit of quantification (<.1 ng/ml) after 168 hours (7 days) (Figure 6C). Everolimus in the myocardium subjacent to scaffold implanted arteries was.71 ng/mg and was measured at 3 hours postimplantation. It quickly decreased to minimally quantifiable levels (.1 ng/mg) by 7 to 14 days. In the lungs, kidneys, spleen, and liver, the drug was only measurable (.1. ng/mg) at 3 hours. Gel Permeation Chromatography The GPC results indicated that the weight-average molecular weight, number-average molecular weight, and polydispersity index of polymer decreased over time. The polymer

7 336 Circ Cardiovasc Interv June months A B C D EEL area (mm ) IEL area (mm ) Lumen area (mm ) Area stenosis (%) Absorb XIENCE V Absorb XIENCE V Absorb XIENCE V Absorb XIENCE V E 3 Injury score F 3 Inflammation score G 1 Struts with giant cells (%) H 3 Fibrin score Absorb XIENCE V Absorb XIENCE V Absorb XIENCE V Absorb XIENCE V Figure 3. Histological analysis for Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb) and XIENCE V. Devices with granulomas (3/1 Absorb [1 at 3 months and at 1 months] and 4/67 XIENCE V [3 at 3 months and 1 at 1 months]) were excluded. A to D, Bar graphs showing external elastic lamina (EEL; A), internal elastic lamina (IEL; B), and lumen areas (C) as well as percent area stenosis (D). Bars represent mean values, and T bars indicate SD. Note late vessel and lumen enlargement observed for Absorb after 1 months. E to H, Box-and-whisker plots showing injury score (E), inflammation score (F), giant cell reaction (G), and fibrin score (H). Lines within boxes represent median values; the upper and lower lines of the boxes represent the 75th and 5th percentiles, respectively; and the upper and lower bars outside the boxes represent the 9th and 1th percentiles, respectively. number-average molecular weight decreased slowly by 18% during the first 6 months of scaffold implantation followed by a more rapid decline thereafter (by 49%, 7%, and 93% at 1, 18, and 4 months, respectively; Figure 7A). The polymer number-average molecular weight dropped to below the measurement limit (4.7 kda) at 3 months after implantation. The percent mass loss of polymer increased slowly by 1% during the first 18 months, followed by a rapid increase thereafter (by 31%, 74%, and 97% at 4, 3, and 36 months, respectively), becoming undetectable at 4 months (limit of detection,.1 mg/ml of polymer, ie, 1.9% of the gravimetric weight of the stent at T [5.3 mg]; Figure 7B). In the 36-month chromatograms, only a small peak below the limit of quantification of polymer (.3 mg/ml) was observed (Figure 7C), indicating the polymer is fully resorbed at 36 months after implantation. Discussion The principal findings in the current preclinical study are as follows: (1) Absorb shows comparable safety to XV in porcine coronary arteries at all time points with patent stents/scaffold and near (1 month) to complete (>1 month) sequestration of struts within the neointima; () mild to moderate inflammation was observed in Absorb which was greater as compared with XV at 6 to 36 months, although inflammation decreased with time; (3) percent area stenosis was greater in Absorb than in XV at all time points except at 3 months, whereas late vessel and lumen enlargement was observed for Absorbimplanted arteries after 1 months; (4) histological changes of Absorb dismantling were observed after 1 months, and resorption sites of Absorb were poorly discernible to essentially absent at 4 months; (5) GPC analysis confirmed that degradation of Absorb was complete by 36 months; and (6) pharmacokinetic studies of everolimus release in Absorb was similar to previously published XV data. These findings not only support the long-term safety of Absorb, but also show lumen enlargement overtime with mild to moderate inflammation. The current preclinical study of Absorb demonstrates the unique characteristics of the scaffold with its ability to allow for lumen enlargement, thus distinguishing it from metallic stents. Delayed arterial healing with poor strut coverage has been shown to be the primary substrate responsible for late and very late stent thrombosis after the first-generation sirolimus- and paclitaxel-eluting stent placement. 16 Autopsy studies have demonstrated divergent mechanisms of late and very late stent thrombosis between sirolimus- and paclitaxel-eluting stents, that is, hypersensitivity reaction to the permanent polymer with extensive inflammation consisting of eosinophils and

8 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold 337 Table 4. Histological Analysis 1 mo 3 mo 6 mo 1 mo 18 mo 4 mo 3 mo 36 mo 4 mo P Value* P Value for Interaction EEL, mm Absorb 8.1±.9 8.3±.7 8.3±.8 9.5± ± ± ± ±1.6 1.±1.5 <.1 <.1 XIENCE V 7.9±1. 9.3±. 8.5±1. 8.7±1. 8.4±.6 9.±.9 9.5±1. 8.3± ±.7.13 P value IEL, mm Absorb 6.7±.8 6.7±.5 7.±.6 8.1± ± ± ± ± ±1.4 <.1 <.1 XIENCE V 6.8±.9 7.1±.6 7.4±.8 7.5±.9 7.4±.6 7.7±.8 8.4±.8 7.6±1. 7.±.7.1 P value Media, mm Absorb 1.35±. 1.56± ±. 1.43± ± ± ±.3.85±.3 1.3± XIENCE V 1.1±.8.5±.8 1.5±. 1.±.38 1.± ± ±.1.7±.8 1.9± P value Lumen, mm Absorb 4.9±.6 4.3±.9 4.7±.7 5.6±1.1 7.± ±1.4 8.± ± ±1.5 <.1. XIENCE V 5.9±.9 4.8± ±.8 6.±.9 6.3±.8 6.5±1. 7.±1. 6.1±1. 6.1±.6.15 P value % Stenosis Absorb 7.9± ± ±6.1 9.± ± ±5. 7.± ± ± XIENCE V 13.5± ±. 3.8± ± ± ± ± ± ±1.6.8 P value Neointimal thickness, mm Absorb.1±.4.18±.9.15±.5.14±.8.13±.4.15±.3.14±.4.16±.6.15± XIENCE V.7±.4.7±.3.14±.3.8±.5.6±.4.7±.5.6±.7.7±.5.4±.1.6 P value Injury score Absorb.9 (.15.44).9 (..18).5 (.14.4).19 (.8.63).44 (.3.95).35 (.9.74).35 (.11.69).68 ( ).63 (..8).3.34 XIENCE V.5 (..1).8 (.1 1.3).14 (.11.).5 (.3).18 ( ).34 (.14.46).19 (.5.6).3 ( ).3 (.8.83).4 P value (Continued )

9 338 Circ Cardiovasc Interv June 14 Table 4. Continued 1 mo 3 mo 6 mo 1 mo 18 mo 4 mo 3 mo 36 mo 4 mo P Value* P Value for Interaction Inflammation score Absorb ( ) (.33).17 (.75) 1. ( ) 1. ( ).67 (.33.9).83 ( ).67 ( ) (.33) <.1.13 XIENCE V ( ) ( 4.) ( ) ( ) (.33) (.33) (.5) (.33) ( ).17 P value Strut with giant cells, % Absorb 35.1 ( ) 1.3 ( 6.6) (.6).6 ( 13.3) 9.3 ( ) ( 6.8) 3.5 (.6 8.7) 1. ( 6.6) ( ) <.1.4 XIENCE V 17.6 ( ) 6.7 (.9 3.1) ( ) (.) 4. ( 16.6) (.8) 1.1 ( 4.9) 1.1 ( 11.7) ( ).31 P value Fibrin score Absorb 1.67 (1.5.).5 (.5 1.) (.17) (.33) (.33) ( ) (.5) ( ) ( ).6.3 XIENCE V 1.5 (1..) (.33) (.33) (.33) ( ) ( ) ( ) ( ) ( ).4 P value The analyses include devices with granulomas which were observed in 3 of 1 Absorb (.9%; 1 at 3 months and at 1 months) and 4 of 67 XIENCE V (6.%; 3 at 3 months and 1 at 1 months). Eosinophil infiltration was absent in both devices except in stents/scaffolds that have a granulomatous reaction, and in these, the number of eosinophils per high-power field varied from to 5 cells which were observed along with lymphocytes and macrophages. Values are expressed as mean±sd or median (5th 75th percentile). Absorb indicates Absorb everolimus-eluting bioresorbable vascular scaffold; EEL, external elastic lamina; and IEL, internal elastic lamina. *P value for the comparison across all time points. P value for interaction with time and device. P value for Absorb vs XIENCE V at each time point.

10 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold 339 A B C D Figure 4. Discontinuities of scaffolds observed in arteries implanted with Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb). A and B, Histological images showing strut irregularity (arrows) and overlaid struts (black arrowhead) observed in the left anterior descending coronary artery (LAD) with Absorb at 1 month. C and D, Histological images showing strut irregularity (arrow) in the LAD with Absorb at 3 months with overlaid struts (black arrowheads) and atypical strut alignment (white arrowhead). Note complete neointimal coverage of all struts in both arteries with increased neointimal growth; however, in both arteries, the lumens remained patent. lymphocytes in the former and malapposition with excessive fibrin deposition in the latter. 1 However, the second-generation XV stents are safer as compared with the first-generation DESs in both clinical and pathological studies, where XV exhibit lower frequency of late and very late stent thrombosis with a strikingly lower prevalence of uncovered struts, less inflammation, and less fibrin deposition versus sirolimus- and paclitaxel-eluting stents. 8,17,18 The current preclinical study showed that Absorb exhibited comparable safety to XV, with near (1 month) to complete (>1 month) sequestration of struts within neointima without evidence of thrombosis. Vascular inflammation plays a crucial role in device outcomes because it has been shown to be associated with Absorb 1 month 3 months 1 months neointimal growth and restenosis. The current study showed mild to moderate inflammation in Absorb through 4 months. However, Absorb exhibited greater inflammation scores as compared with XV from 6 to 36 months. Nevertheless, the inflammation score, although greater in Absorb, was generally <1 except at 1, 18, 3, and 36 months with 75th percentile reaching >1. Greater inflammation has been correlated with greater neointimal thickness and area stenosis in bare metal stents and DESs. 19, It has been reported that the rate of degradation affects the degree of inflammatory response. 1 Thus, it may be hypothesized that inflammation associated with Absorb is related to polymer resorption. However, the observed inflammation XIENCE V 1 month 3 months 1 months Figure 5. Representative scanning electron microscopy images of Absorb bioresorbable vascular scaffold (Absorb) and XIENCE V at 1, 3, and 1 months after device placement in porcine coronary arteries. The low power images acquired at 15 magnification show complete strut coverage by mildly thickened neointima at 1 month in both Absorb and XIENCE V, which are maintained at 3 and 1 months. Highpower images acquired at 6 magnification show completely endothelialized luminal surface with cells arranged in cobblestone morphology and showing well-formed cell-to-cell contacts that indicate the mature nature of the endothelium in both Absorb and XIENCE V at 1, 3, and 1 months. Both devices show infrequent adherent inflammatory cells and scattered red blood cells on luminal surfaces with no evidence of luminal thrombus.

11 34 Circ Cardiovasc Interv June 14 A Percent Drug Released B Drug Concentration in Scaffolded Artery Segment (ng/mg) C Everolimus in Blood (ng/ml) 1% 8% 6% 4% % % Time (Days) Mean ±SD n = 6 or Time (Days) Time (Hours) Mean ±SD n = 6 or 7 Mean ±SD n = 3 Figure 6. Pharmacokinetics showing drug release profile (A), drug concentrations in implanted vessel segment (B), and drug concentrations in whole blood (C) after Absorb implantation. A, The mean cumulative percentage of everolimus released during the first 8 days was 79% and drug release peaked at 9 days (96%). B, Everolimus concentration in scaffolded arterial segments was the greatest at 3 hours after implantation with a mean concentration being 16. ng/mg. Arterial everolimus concentrations ranged from.3 to 4.6 ng/mg in 1 to 8 days, which were greater than IC 5 of smooth muscle cell (SMC) that is required for the inhibition of human SMC proliferation. (C) The concentration of everolimus in blood peaked at 15 minutes after Absorb implantation with a mean value being 3.75 ng/ml and quickly declined to below the limit of quantification (<.1 ng/ml) after 168 hours (7 days). progressively decreased after 18 months, which is the same period during which the most rapid mass loss occurred. Macrophages, which feature diverse and plastic phenotypes, influence endothelial proliferation and angiogenesis together with SMC infiltration and collagen deposition, all of which are required for healing and tissue remodeling. At 4 months, there was absence to minimal inflammation when the scaffold is largely resorbed and replaced by matrix consisting of collagen and proteoglycan with infrequent spindle A Mn (kda) B Mass Loss (%) C mv Limit of Measurement Time (months) Time (months) Minutes Figure 7. Gel permeation chromatography (GPC) for the assessment of degradation of Absorb everolimus-eluting bioresorbable vascular scaffold (Absorb). A, The in vivo degradation of polymer of Absorb over time. The lower limit of measurement for the number-average molecular weight (Mn) is 4.7 kda. B, The in vivo mass loss of polymer of Absorb over time. The limit of detection for the mass loss is.1 mg/ml (.1 mg of average initial mass of 5. mg, equivalent to 1.9% of initial mass), and the limit of quantification is.3 mg/ml (.3 mg of average initial mass of 5. mg, equivalent to 5.8% of initial mass). The polymer mass loss values at 1 and 1 months were not reported, because they were likely overestimated because of incomplete polymer recovery. C (Top to bottom), Representative chromatograms of polymer samples before Absorb implantation and at 18, 4, 3, 36, and 4 months after implantation and of tissue blank. Peaks 1 to 5 represent poly(l-lactide) polymer peaks obtained before Absorb implantation and at 18, 4, 3, and 36 months, respectively, after scaffold implantation. The shift of these peaks to the right (lower weightaverage molecular weight range) relative to the polymer peak on the chromatogram from samples at T are consistent with polymer degradation. *Trace tissue species. Everolimus peak

12 Otsuka et al Preclinical Safety of a Bioresorbable Scaffold 341 cells; no inflammatory cells were observed in these regions. The absence or minimal inflammation with largely restored morphological appearance of Absorb-implanted arteries at 4 months indicates that there were no untoward effects of degradation, because the inflammatory response was controlled because of the gradual degradation of the scaffold during a 3-year period. Moreover, the restored morphological appearance of arteries along with the observed expansile remodeling suggests that the inflammation in Absorb-implanted arteries at 6 to 36 months precedes the arterial repair and remodeling. 3 Fibrin deposition is considered to be one of the biological signs of antiproliferative drug effects. The current study showed similar extent of fibrin deposition between Absorb and XV, which peaked at 1 month and rapidly decreased at 3 months followed by absent or minimal deposition at 6 months. Comparable healing processes with decreasing fibrin deposition in Absorb and XV are attributable to similarity in drug release kinetics between the devices as shown in the current pharmacokinetics profile. It is also known that fibrin degradation products can induce SMC migration and proliferation. 4 However, because of less release of the drug after 1 month, it is likely that fibrin deposition is replaced by neointimal tissue in both Absorb and XV, which is supported by similar response cures with peak neointimal growth at 3 to 6 months with regression thereafter attributable to replacement of proteoglycans by collagen followed by cross-linking of collagen that results in a reduction in the neointimal thickness. The neointimal thickness and percent area stenosis were almost always greater in Absorb as compared with XV, which might be at least partly attributable to thicker strut thickness in Absorb (156 μm) versus XV (97 μm). The neointimal thickness in Absorb did not decrease with time but remained similar after 1 months, whereas in XV, the neointima decreased at 1 months. Nevertheless, the lumen area, which is the source of oxygenation and other nutrients to the myocardial bed, was significantly greater in Absorb than XV at 36 and 4 months. The drawbacks (increased stenosis and greater neointimal thickness) of the Absorb scaffold seen at early time points were no longer pertinent at 36 months. These changes can be equated to those observed in nonsignificant atherosclerotic stenosis ( 4%) observed in humans that does not result in lumen compromise. 5 The current study also showed a lower balloon-to-artery ratio in Absorb as compared with XV, whereas injury score was generally greater in Absorb than XV where the difference was significant at 1 months. These results are consistent with the fact that the balloon-to-artery ratio does not always correspond to injury scores. 6 Nevertheless, it is possible that because of greater thickness and radiolucency of the scaffold strut, the balloon-to-artery ratio would be less. There were Absorbs with histological observations suggestive of premature scaffold fracture. Although not specifically known, the inflation rates used during implantation of these scaffolds may have exceeded those detailed in the instruction for use for Absorb ( atm per 5 seconds). Although these observations impacted the degree of lumen narrowing in both arteries, the arterial lumens remained patent, and there were no adverse responses in the distal myocardial beds. GPC analysis parallels histological observations in that minimal mass loss and change in the scaffold s histological appearance are observed through the first 1 months. Morphological changes in the scaffold resorption sites began at 18 months but were most readily appreciated at 4 months and thereafter, which corresponds to the time when numberaverage molecular weight approaches and then falls below the limit of measurement. Similarly, dissolution of resorption sites with connective tissue integration is readily appreciated at 36 and 4 months when the polymer mass is at or below the limit of quantification. The vessel morphology at 4 months of Absorb-implanted arteries showed an intact media with mild neointimal thickening. As anatomy follows function, these anatomic features suggest that Absorb is likely to return the vessel to its original physiological state and reactivity although with mild intimal thickening and medial thinning as compared with the native artery. Study Limitations Vascular responses to the devices in healthy swine coronary arteries are likely different from those in diseased human arteries, which may limit the ability to generalize the current findings directly to clinical outcomes in humans. Inflammatory reaction may be different because no underlying inflammation was present at the time of implantation in the porcine model. The current study used different species of swine (Yorkshire for 1 3 months and Yucatan for 6 months) and different lengths of stent/scaffold (3. 18 mm for 1 6 months and 3. 1 mm for 1 months), which may have affected the vascular response to the implants. Also, Yorkshire swine have a tendency to elicit a greater inflammatory response than do Yucatan swine; however, no adverse effects were ascertained throughout the continuum of the study. It should also be noted that angiographic assessment has limited accuracy as compared with histological analysis; the latter is more reflective of the actual morphometric changes over time. SEM assessment was performed in only limited number of devices in normal arteries. Moreover, the multiple hypotheses testing attributable to the many time points could result in false statistical significance because of chance. Nevertheless, our study shows a definite safety profile of Absorb with comparable results to XV, the latter of which has been in clinical use for >8 years. Conclusions The Absorb bioresorbable vascular scaffold demonstrated comparable safety to XV in the normal porcine coronary arteries 4 months after implantation with mild to moderate inflammation and greater neointimal thickening, along with the unique response of late lumen and vessel enlargement seen at 36 months. Pharmacokinetic studies of release of everolimus were similar between Absorb and XV and were associated with fibrin deposition in both at 1 month. Degradation in Absorb was completed within 36 months in GPC analysis. Nevertheless, Absorb needs to be studied in large clinical trials which are currently ongoing. Acknowledgments We thank Leslie Afan (Abbott Vascular) for her technical support with data analysis, archiving, and retrieval; Katherine Fu and David

13 34 Circ Cardiovasc Interv June 14 Pinson (Abbott Vascular) for their technical assistance with GPC analysis; and Byron Lambert, Alexander Sheehy, Daniel Cox, and James Benham (Abbott Vascular) for their scientific input on the article. We also acknowledge American Preclinical Services for their care and dedication in the successful completion of the in-life phases and angiographic data acquisition for this study. Sources of Funding This study was sponsored by Abbott Vascular. CVPath Institute, Inc, a private nonprofit research organization, provided partial support for this work. Dr Otsuka is supported by a research fellowship from the Uehara Memorial Foundation, Tokyo, Japan. Disclosures Dr Virmani receives research support from Abbott Vascular, BioSensors International, Biotronik, Boston Scientific, Medtronic, MicroPort Medical, OrbusNeich Medical, SINO Medical Technology, and Terumo Corporation; has speaking engagements with Merck; receives honoraria from Abbott Vascular, Boston Scientific, Lutonix, Medtronic, and Terumo Corporation; and is a consultant for 48 Biomedical, Abbott Vascular, Medtronic, and W.L. Gore. Dr Perkins, J.P. Lane, Dr Q. Wang, Dr Kamberi, Dr J. Wang, and Dr Rapoza are employees of Abbott Vascular. Dr Sakakura has received speaking honorarium from Abbott Vascular, Boston Scientific, and Medtronic CardioVascular. The other authors report no conflicts. References 1. Nakazawa G, Finn AV, Vorpahl M, Ladich ER, Kolodgie FD, Virmani R. Coronary responses and differential mechanisms of late stent thrombosis attributed to first-generation sirolimus- and paclitaxel-eluting stents. J Am Coll Cardiol. 11;57: Nakazawa G, Otsuka F, Nakano M, Vorpahl M, Yazdani SK, Ladich E, Kolodgie FD, Finn AV, Virmani R. The pathology of neoatherosclerosis in human coronary implants bare-metal and drug-eluting stents. J Am Coll Cardiol. 11;57: Serruys PW, Garcia-Garcia HM, Onuma Y. From metallic cages to transient bioresorbable scaffolds: change in paradigm of coronary revascularization in the upcoming decade? Eur Heart J. 1;33:16 5b. 4. Ormiston JA, Serruys PW, Regar E, Dudek D, Thuesen L, Webster MW, Onuma Y, Garcia-Garcia HM, McGreevy R, Veldhof S. A bioabsorbable everolimus-eluting coronary stent system for patients with single de-novo coronary artery lesions (ABSORB): a prospective open-label trial. Lancet. 8;371: Serruys PW, Ormiston JA, Onuma Y, Regar E, Gonzalo N, Garcia-Garcia HM, Nieman K, Bruining N, Dorange C, Miquel-Hébert K, Veldhof S, Webster M, Thuesen L, Dudek D. A bioabsorbable everolimus-eluting coronary stent system (ABSORB): -year outcomes and results from multiple imaging methods. Lancet. 9;373: Ormiston JA, Serruys PW, Onuma Y, van Geuns RJ, de Bruyne B, Dudek D, Thuesen L, Smits PC, Chevalier B, McClean D, Koolen J, Windecker S, Whitbourn R, Meredith I, Dorange C, Veldhof S, Hebert KM, Rapoza R, Garcia-Garcia HM. First serial assessment at 6 months and years of the second generation of absorb everolimus-eluting bioresorbable vascular scaffold: a multi-imaging modality study. Circ Cardiovasc Interv. 1;5: Stone GW, Rizvi A, Sudhir K, Newman W, Applegate RJ, Cannon LA, Maddux JT, Cutlip DE, Simonton CA, Sood P, Kereiakes DJ; SPIRIT IV Investigators. Randomized comparison of everolimus- and paclitaxeleluting stents. -year follow-up from the SPIRIT (Clinical Evaluation of the XIENCE V Everolimus Eluting Coronary Stent System) IV trial. J Am Coll Cardiol. 11;58: Otsuka F, Vorpahl M, Nakano M, Foerst J, Newell JB, Sakakura K, Kutys R, Ladich E, Finn AV, Kolodgie FD, Virmani R. Pathology of secondgeneration everolimus-eluting stents versus first-generation sirolimus- and paclitaxel-eluting stents in humans. Circulation. 14;19: Perkins LEL, Boeke-Purkis KH, Wang Q, Stringer S, Coleman LA. Xience v tm everolimus-eluting coronary stent system: A preclinical assessment. J Interven Cardiol. 9;:S8 S4. 1. Nakazawa G, Nakano M, Otsuka F, Wilcox JN, Melder R, Pruitt S, Kolodgie FD, Virmani R. Evaluation of polymer-based comparator drugeluting stents using a rabbit model of iliac artery atherosclerosis. Circ Cardiovasc Interv. 11;4: Lane J, Perkins L, Sheehy A, Pacheco E, Frie M, Lambert B, Rapoza R, Virmani R. Lumen gain and restoration of pulsatility following implantation of a bioresorbable vascular scaffold in porcine coronary arteries [published online ahead of print May 9, 14]. JACC Cardiovasc Interv. doi: 1.116/j.jcin Schwartz RS, Huber KC, Murphy JG, Edwards WD, Camrud AR, Vlietstra RE, Holmes DR. Restenosis and the proportional neointimal response to coronary artery injury: results in a porcine model. J Am Coll Cardiol. 199;19: Finn AV, Kolodgie FD, Harnek J, Guerrero LJ, Acampado E, Tefera K, Skorija K, Weber DK, Gold HK, Virmani R. Differential response of delayed healing and persistent inflammation at sites of overlapping sirolimus- or paclitaxel-eluting stents. Circulation. 5;11: Schuler W, Sedrani R, Cottens S, Häberlin B, Schulz M, Schuurman HJ, Zenke G, Zerwes HG, Schreier MH. SDZ RAD, a new rapamycin derivative: pharmacological properties in vitro and in vivo. Transplantation. 1997;64: Kovarik JM, Kaplan B, Tedesco Silva H, Kahan BD, Dantal J, Vitko S, Boger R, Rordorf C. Exposure-response relationships for everolimus in de novo kidney transplantation: defining a therapeutic range. Transplantation. ;73: Joner M, Finn AV, Farb A, Mont EK, Kolodgie FD, Ladich E, Kutys R, Skorija K, Gold HK, Virmani R. Pathology of drug-eluting stents in humans: delayed healing and late thrombotic risk. J Am Coll Cardiol. 6;48: Palmerini T, Biondi-Zoccai G, Della Riva D, Stettler C, Sangiorgi D, D Ascenzo F, Kimura T, Briguori C, Sabatè M, Kim HS, De Waha A, Kedhi E, Smits PC, Kaiser C, Sardella G, Marullo A, Kirtane AJ, Leon MB, Stone GW. Stent thrombosis with drug-eluting and bare-metal stents: evidence from a comprehensive network meta-analysis. Lancet. 1;379: Räber L, Magro M, Stefanini GG, Kalesan B, van Domburg RT, Onuma Y, Wenaweser P, Daemen J, Meier B, Jüni P, Serruys PW, Windecker S. Very late coronary stent thrombosis of a newer-generation everolimus-eluting stent compared with early-generation drug-eluting stents: a prospective cohort study. Circulation. 1;15: Farb A, Weber DK, Kolodgie FD, Burke AP, Virmani R. Morphological predictors of restenosis after coronary stenting in humans. Circulation. ;15: Wilson GJ, Nakazawa G, Schwartz RS, Huibregtse B, Poff B, Herbst TJ, Baim DS, Virmani R. Comparison of inflammatory response after implantation of sirolimus- and paclitaxel-eluting stents in porcine coronary arteries. Circulation. 9;1:141 9, Shive MS, Anderson JM. Biodegradation and biocompatibility of PLA and PLGA microspheres. Adv Drug Deliv Rev. 1997;8:5 4.. Brown BN, Londono R, Tottey S, Zhang L, Kukla KA, Wolf MT, Daly KA, Reing JE, Badylak SF. Macrophage phenotype as a predictor of constructive remodeling following the implantation of biologically derived surgical mesh materials. Acta Biomater. 1;8: Burke AP, Kolodgie FD, Farb A, Weber D, Virmani R. Morphological predictors of arterial remodeling in coronary atherosclerosis. Circulation. ;15: Naito M, Stirk CM, Smith EB, Thompson WD. Smooth muscle cell outgrowth stimulated by fibrin degradation products. The potential role of fibrin fragment E in restenosis and atherogenesis. Thromb Res. ;98: Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med. 1987;316: Kornowski R, Hong MK, Tio FO, Bramwell O, Wu H, Leon MB. In-stent restenosis: contributions of inflammatory responses and arterial injury to neointimal hyperplasia. J Am Coll Cardiol. 1998;31:4 3.