Aai

AB

ARRAA

KCDBS

1

brAst6aastc(b

0h

Applied Surface Science 279 (2013) 216– 232

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

jou rn al h omepa g e: www.elsev ier .com/ locate /apsusc

stent for co-delivering paclitaxel and nitric oxide from abluminalnd luminal surfaces: Preparation, surface characterization, andn vitro drug release studies

nnemarie Gallo, Gopinath Mani ∗

iomedical Engineering Program, The University of South Dakota, Sioux Falls, SD 57107, USA

a r t i c l e i n f o

rticle history:eceived 21 February 2013eceived in revised form 16 April 2013ccepted 18 April 2013vailable online 28 April 2013

eywords:ardiovascular stentrug-eluting stentsiomaterials surface modificationurface characterization

a b s t r a c t

Most drug-eluting stents currently available are coated with anti-proliferative drugs on both abluminal(toward blood vessel wall) and luminal (toward lumen) surfaces to prevent neointimal hyperplasia. Whilethe abluminal delivery of anti-proliferative drugs is useful for controlling neointimal hyperplasia, theluminal delivery of such drugs impairs or prevents endothelialization which causes late stent thrombosis.This research is focused on developing a bidirectional dual drug-eluting stent to co-deliver an anti-proliferative agent (paclitaxel – PAT) and an endothelial cell promoting agent (nitric oxide – NO) fromabluminal and luminal surfaces of the stent, respectively. Phosphonoacetic acid, a polymer-free drugdelivery platform, was initially coated on the stents. Then, the PAT and NO donor drugs were co-coated onthe abluminal and luminal stent surfaces, respectively. The co-coating of drugs was collectively confirmedby the surface characterization techniques such as Fourier transform infrared spectroscopy, scanningelectron microscopy (SEM), 3D optical surface profilometry, and contact angle goniometry. SEM showedthat the integrity of the co-coating of drugs was maintained without delamination or cracks formation

occurring during the stent expansion experiments. In vitro drug release studies showed that the PAT wasreleased from the abluminal stent surfaces in a biphasic manner, which is an initial burst followed by aslow and sustained release. The NO was burst released from the luminal stent surfaces. Thus, this studydemonstrated the co-delivery of PAT and NO from abluminal and luminal stent surfaces, respectively.The stent developed in this study has potential applications in inhibiting neointimal hyperplasia as wellas encouraging luminal endothelialization to prevent late stent thrombosis.

. Introduction

Coronary artery disease (CAD) is the leading cause of death foroth men and women throughout the world [1]. CAD is the nar-owing of blood vessels that supply blood and oxygen to the heart.

balloon angioplasty is carried out to open up the narrowed arteryo that the blood flow to the heart can be restored [2,3]. However,he artery renarrows (restenosis) in 30–40% of the patients within

months after balloon angioplasty [4,5]. Currently, metallic stentsre implanted followed by angioplasty to keep the artery open for

longer period of time. The implantation of bare metal stents hasignificantly reduced the restenosis rate to 20–30% [4,5]. However,he arterial injury that occurs during stent implantation causes a

ascade of biological events resulting in neointimal hyperplasiaNH) [6]. NH is a natural wound healing response characterizedy the growth and migration of smooth muscle cells followed by

∗ Corresponding author. Tel.: +1 605 367 7773; fax: +1 605 782 3280.E-mail addresses: Gopinath.Mani@usd.edu, mgopitex@hotmail.com (G. Mani).

169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2013.04.072

© 2013 Elsevier B.V. All rights reserved.

the deposition of extracellular matrix inside the arterial lumen [7].This results in re-occlusion of artery even after the implantation ofstents, which is called in-stent restenosis. Drug-eluting stents (DES)are currently implanted to treat NH by releasing anti-proliferativedrugs which can inhibit the growth of smooth muscle cells insidethe lumen [8,9]. Although the restenosis rate has been reduced to<10% after the implantation of DES, the occurrence of late stentthrombosis (LST) is the major safety concern of DES at present [10].LST is an adverse clinical event characterized by the formation ofblood clots in the arteries after months or years of DES implan-tation and results in heart attack or death [11–13]. The primarycause of LST is believed to be the delayed or impaired endothelialcell growth on DES [11–13].

Most DES release anti-proliferative drugs in abluminal (towardblood vessel wall) as well as luminal (toward lumen – the inneropen space or cavity of a blood vessel is called lumen) directions

[10,14]. While the abluminal release of anti-proliferative drugsis highly beneficial in controlling the growth of smooth musclecells and thereby inhibiting neointimal hyperplasia, the luminalrelease of such drugs impedes re-endothelialization [12,15–17].

rface Science 279 (2013) 216– 232 217

Titaptedcls

saauppndnsitAalht

(pPcdmdritd

2

2

c1(m(2wN((nCC

2

e

A. Gallo, G. Mani / Applied Su

he re-endothelialization of luminal stent surfaces is of paramountmportance because the complete endothelial cell lining preventshe adhesion, aggregation, and activation of blood platelets [18]nd thereby inhibits LST. Hence, there is a need to co-deliver thera-eutic agents from abluminal and luminal stent surfaces to inhibithe growth of smooth muscle cells and to encourage the growth ofndothelial cells, respectively. The research goal of this study is toevelop a bidirectional dual drug-eluting coronary stent which cano-deliver an antiproliferative agent (paclitaxel) and an endothe-ial cell promoting agent (nitric oxide) from abluminal and luminalurfaces, respectively.

Polymeric carriers are commonly used to deliver drugs fromtents. Some (not all) polymers used in DES can cause inflammatorynd hypersensitive reactions [19–22]. Hence, the research in therea of drug delivery platforms for stents can be broadly classifiednder the following two categories: (a) developing more biocom-atible polymer platforms; (b) developing totally polymer-freelatforms. In this study, a polymer-free platform using phospho-oacetic acid based molecular coatings was used to co-deliver therugs from stents. The reasons for choosing paclitaxel (PAT) anditric oxide (NO) to be co-delivered from abluminal and luminaltent surfaces, respectively are as follows. The effect of PAT onnhibiting the growth of smooth muscle cells and preventing neoin-imal hyperplasia has been well reported in the literature [23–25].lso, PAT has been used in U.S. Food and Drug Administration (FDA)pproved stents [25]. NO is well known for enhancing endothe-ialization of different material surfaces [26–28]. In addition, NOas been shown to inhibit neointimal hyperplasia [27,29,30] andhrombosis [26,27,31].

In this study, five different groups of stents were used. They area) chemically cleaned control stents; (b) stents coated with onlyhosphonoacetic acid with no drugs on either side; (c) stents withAT coating on the abluminal surface and no nitric oxide donor drugoating on the luminal surface; (d) stents with a nitric oxide donorrug coating on the luminal surface and no PAT coating on the ablu-inal surface; (e) stents co-coated with PAT and nitric oxide donor

rug on the abluminal and luminal stent surfaces, respectively. Theeasons for using the single drug coated stents (groups – c and d)n this study are that these stents are not only the right controls forhe co-coated stent but also to make sure that no differences ariseuring the co-coating process.

. Materials and methods

.1. Materials

Cobalt–chromium (Co–Cr) alloy bare metal stents were pur-hased from Fortimedix B.V. (Netherlands). These stents were1.2 mm in length with strut dimensions of 0.083 mm × 0.092 mmwidth × thickness). Absolute ethanol (200 proof), acetone,

ethanol, dimethyl sulfoxide (DMSO), phosphonoacetic acidPAA), and phosphate-buffered saline with 0.05% tween-20 (PBS/T-0, pH = 7.4) were obtained from Sigma–Aldrich (USA). HPLC-gradeater and acetonitrile were also purchased from Sigma–Aldrich.itric oxide donor drug diethylenetriamine diazeniumdiolate

DETA NONOate; formula name: (Z)-1-[N-(2-aminoethyl)-N-2-ammonioethyl)amino]diazen-1-ium-1,2-diolate) and nitrate/itrite colorimetric assay kit were purchased from Caymanhemical (Ann Arbor, MI). Paclitaxel (PAT) was purchased fromhemieTek (Indianapolis, IN). All chemicals were used as received.

.2. Chemical cleaning of stents

Co–Cr alloy stents were chemically cleaned by sonicating inthanol, acetone, and methanol twice for 10 min each with fresh

Fig. 1. Chemical structure of (A) phosphonoacetic acid (PAA), (B) paclitaxel(PAT) and (C) nitric oxide donor drug diethylenetriamine diazeniumdiolate (DETANONOate).

solvents used each time. The stents were then dried using N2 gas.The chemically cleaned stents are referred to here as control stents# 1 (CS#1).

2.3. Phosphonoacetic acid coating on stents

The chemically cleaned stents were immersed in 3 mL of 1 mMsolution of phosphonoacetic acid (PAA, chemical structure shownin Fig. 1A) in de-ionized water (DI-H2O) for 24 h. The stents werethen taken out of the solution and transferred to an oven withoutrinsing. In the oven, the stents were heat treated in air at 120 ◦Cfor 18 h. The stents were then cleaned by sonication in DI-H2O for1 min followed by N2 gas drying. Thus prepared PAA coated stentsare referred to here as control stents # 2 (CS#2).

2.4. Paclitaxel coating on the abluminal surface of the stent

Paclitaxel (PAT, chemical structure shown in Fig. 1B) was spraycoated on the stents. For coating PAT on the abluminal stent surface,initially, the stent was placed on a mandrel in which the lumi-nal stent surface was in close contact with the mandrel. This wasdone to prevent any PAT leaking onto the luminal stent surface dur-

ing spray coating. A solution of PAT at a concentration of 1 mg/mLwas prepared in a solvent mixture containing 75% ethanol and 25%DMSO. The PAT solution was sprayed on the stents using a Medi-Coat stent coating system (Sono-Tek Corporation, Milton, NY). The

218 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

Table 1Parameters used for spray coating stents.

Parameters Values

Distance from nozzle tip to stent 0.35 in.Ultrasonic power 0.7 WSyringe pump dispense rate 0.02 mL/minFocusing gas pressure 0.5 psiDrying gas pressure 2 psiRotation rate 70 rpmHorizontal translation speed 0.1 in./s (2.5 mm/s)Number of loops 120 loops

psldrwPltPh

2l

FDft5cft1s

2l

ctoiarFt

2

isc

2

Fas

DETA NONOate coating on the luminal stent surface, the stent wasinitially cut in the longitudinal direction. Then, the stent was lon-gitudinally opened and carefully placed in the ATR with only the

Table 2Abbreviations used and their descriptions.

Abbreviations Descriptions

PAA Phosphonoacetic acidPAT PaclitaxelDETA NONOate Diethylenetriamine diazeniumdiolate nitric oxide donor

drugNO Nitric oxideCS#1 Control stent # 1: Chemically cleaned stentCS#2 Control stent # 2: PAA coated stentABL-PAT PAT coated on the abluminal stent surface (no drug coated

on the luminal stent surface)

Drying time in nitrogen 30 minDrying time in air 15 min

arameters used for spray coating are provided in Table 1. Afterpray coating, the stent was removed from the mandrel and theuminal stent surface alone was cleaned using the following proce-ure. A clean mandrel was dipped in fresh ethanol for 10 s. Uponemoval, the PAT coated stent was carefully placed onto the ethanoletted mandrel and slowly moved back and forth to remove any

AT, if present on the luminal stent surface. This mandrel-baseduminal stent cleaning procedure was repeated five times to ensurehat no PAT was present on the luminal stent surface. Thus preparedAT coated (only on the abluminal surface) stents are referred toere as ABL-PAT stents.

.5. Nitric oxide donor drug DETA NONOate coating on theuminal stent surface

A 5 mM solution of DETA NONOate (chemical structure shown inig. 1C) was prepared in DI-H2O. A clean mandrel was wetted withETA NONOate by immersing it in 3 mL of the prepared solution

or 30 min. As soon as the mandrel was removed from the solution,he stent to be luminally coated was placed on the mandrel for

min which allowed only the luminal stent surface to be in closeontact with DETA NONOate wetted mandrel. This process trans-erred DETA NONOate from the mandrel to the luminal surface ofhe stent. The luminally coated stents were then dried in air for5 min. Thus prepared DETA NONOate coated (only on the luminalurface) stents are referred to here as LUM-NO stents.

.6. Co-coating of PAT and DETA NONOate on the abluminal anduminal surfaces of the stent, respectively

The co-coating of PAT and DETA NONOate was carried out byombining the procedures described in Sections 2.4 and 2.5. Briefly,he PAA coated stent (Section 2.3) was spray coated with PAT onlyn the abluminal surface (Section 2.4) followed by the luminal coat-ng of DETA NONOate (Section 2.5). Thus prepared co-coated stentsre referred to here as Co-PAT-NO stents. A schematic of the prepa-ation of PAT and DETA NONOate co-coated stents is provided inig. 2. Also, the different abbreviations used in this manuscript andheir descriptions are provided in Table 2.

.7. Surface characterization

All stents (CS#1, CS#2, ABL-PAT, LUM-NO, and Co-PAT-NO) usedn the study were characterized using Fourier transform infraredpectroscopy (FTIR), scanning electron microscopy (SEM), 3D opti-al surface profilometry, and contact angle goniometry.

.7.1. Fourier-transform infrared spectroscopy (FTIR)

The infrared spectra (IR) were collected using a Nicolet 6700

TIR spectrometer (Thermo Scientific, USA) equipped with anttenuated total reflection (ATR) smart performer accessory. Eachpectrum was collected with 32 or 64 scans at a spectral resolution

Fig. 2. Schematic of preparation of paclitaxel and DETA NONOate co-coated coro-nary stent.

of 4 cm−1. All the IR spectra presented here were baseline correctedand analyzed using OMNICS software.

To collect IR spectrum of the PAT coating on the abluminal sur-face, the stent was carefully placed on the ZnSe crystal of the ATRaccessory with only the abluminal surface touching the crystal.A clamp was placed on the stent to make the abluminal surfacein close contact with the crystal, which was followed by the col-lection of IR spectrum using ATR. To collect IR spectrum of the

LUM-NO DETA NONOate coated on the luminal surface of the stent(no drug coated on the abluminal stent surface)

Co-PAT-NO PAT and DETA NONOate co-coated on the abluminal andluminal stent surfaces, respectively

A. Gallo, G. Mani / Applied Surface S

Table 3Parameters used in 3D optical surface profilometry characterization.

Parameters Values

Magnification 99.5×Measurement array size 640 × 480Lateral sampling 0.1 �mField of view 64 �m × 48 �m3D filter Gaussian – short wave pass

Long wavelength cutoff = 0.01 mmHeight resolution <0.0003 �m (PSI mode)

<0.006 �m (VSI mode)Bearing ratio offsets peak/valley 1%/1%Stylus X �c/�s 60 �m/0.6 �m

lc

2

oeTp

2

Mdpavo

2

tHwwtatdm

2

pfta

2s

swaH

a wavelength of 540 nm using an Infinite M200 microplate reader

Stylus Y �c/�s 35 �m/0.4 �mStylus filter type Gaussian

uminal surface in close contact with the crystal, followed by theollection of IR spectrum.

.7.2. Scanning electron microscopy (SEM)A Quanta 450 SEM from FEI (USA) was used in this study to

btain the images of surface morphology of the stents. An accel-rating voltage of 30 kV was used at a working distance of 10 mm.he stents were sputter coated with a 20 nm thick layer of gold-alladium prior to SEM imaging.

.7.3. 3D optical surface profilometryA Wyko NT8000 optical surface profilometer (operated at

ichigan Metrology LLC) was used in this study to obtain three-imensional (3D) topography images of the stents. The differentarameters that were employed in the optical profilometer char-cterization of stents are provided in Table 3. The RMS roughnessalues reported here represent the average of three distinct spotsn a sample along with the corresponding standard deviation.

.7.4. Contact angle goniometryA VCA optima (AST products, Inc.) system was used in this study

o measure contact angles on the stents. A 0.5 �L volume of DI-2O was placed on the stent surfaces and the static contact anglesere measured on both the sides of the drop after 15 s. The stentsere longitudinally cut and opened (as described in Section 2.7.1)

o measure contact angles on the luminal surfaces while the contactngles were measured on the abluminal surfaces without cuttinghe stents. The contact angles were measured on at least threeistinct spots on a stent and the reported values represent theean ± standard deviation of the measurements.

.8. Expansion of stents using balloon catheter

The control and co-coated stents were mounted on an angio-lasty balloon catheter and were expanded to 3 mm in diameteror 60 s at a pressure of 9.5 atm. The integrity of the co-coating andhe surface morphology of the expanded stents were studied using

SEM.

.9. Quantification of paclitaxel on the abluminal surfaces of thetent

The PAT coated stents (ABL-PAT and Co-PAT-NO stents) wereonicated in 4 mL of ethanol thrice for 10 min each. Fresh solvent

as used each time. The solutions collected each time were then

nalyzed for the quantity of PAT extracted from the stents usingPLC.

cience 279 (2013) 216– 232 219

2.10. Quantification of nitric oxide on the luminal surfaces of thestent

DETA NONOate coated stents (LUM-NO and Co-PAT-NO stents)were sonicated in 1 mL of DI-H2O six times for 10 min each. Freshsolvent was used each time. The solutions collected each time werethen analyzed for the concentration of nitric oxide extracted fromthe stents using a nitrate/nitrite colorimetric assay kit.

2.11. In vitro drug release studies

The drug coated stents (ABL-PAT, LUM-NO, and Co-PAT-NO)were immersed in 2 mL of PBS/Tween-20 (pH = 7.4) and incubatedin a circulating water bath (Thermo Scientific, USA) at 37 ◦C. Tween-20 is a non-ionic surfactant which is commonly added in PBSto increase the solubility of PAT and to maintain sink conditions[32–35]. At pre-determined time points (1 h, 3 h, 6 h, 12 h, and 24 h,and every day thereafter for up to 14 days, followed by day-21 andday-28), the stents were taken out of the solution and transferredto fresh PBS/T-20 solution. The PBS/T-20 solution collected at eachtime point was then analyzed for the quantity of PAT and nitricoxide released using HPLC and nitrate/nitrite colorimetric assay,respectively.

2.11.1. High-performance liquid chromatography (HPLC)The HPLC used in this study was Waters e2695 system equipped

with a Waters 2489 UV/vis detector and a Nova-Pak C18 col-umn (3.9 mm × 150 mm; particle size – 4 �m). The HPLC protocolsused for determining the amount of PAT released were followedas reported previously [36]. Briefly, a mobile phase of water andacetonitrile (45:55, V/V) was used at a flow rate of 1 mL/min. Aninjection volume of 10 �L was used for the analysis. The columntemperature was maintained at 35 ◦C. The UV detector wave-length was set at 227 nm. For the calibration samples, a linearplot with a correlation coefficient of R2 = 0.999 was obtained forthe concentration range of 3 ng/mL to 50 �g/mL. A Waters Mil-lennium 32 software was used to analyze all the collected HPLCdata.

2.11.2. Nitrate/nitrite colorimetric assayThe nitrate/nitrite colorimetric assay used in this study meas-

ures the total nitrate/nitrite concentration in a sample. When thenitric oxide (NO) is generated, it rapidly oxidizes into nitrates andnitrites, which are commonly measured to quantify the total NOproduction [31,37–39]. As a two-step assay, in the first step, thenitrates are converted into nitrites using nitrate reductase. In thesecond step, the nitrites are converted into a deep purple azo com-pound using Griess reagent. The absorbance of the azo compoundis measured photometrically to determine the concentration ofnitrites. Griess reagent based assays are commonly used in the lit-erature to determine the NO production [26,31,39]. The assay wascarried out as per the details provided in the kit. Briefly, a 80 �Laliquot of the sample was taken in a 96 well plate. Then, a 10 �Lof enzyme cofactor mixture was added to the sample followed bythe addition of 10 �L of nitrate reductase mixture. After incubatingthese contents at the room temperature for 1 h, a 50 �L of GriessReagent R1 was added followed by the addition of 50 �L of GriessReagent R2. The contents were then incubated at the room temper-ature for 10 min. The absorbance of the solution was measured at

(Tecan, USA). A linear plot with a correlation coefficient of R2 = 0.99was obtained for the calibration samples ranging from 1 �M to35 �M.

2 rface

2

aTm

3

3

TamIbapatwlvpPtbbpgwwtghutosodntft

mosotwFtNpaotm3t

ns

20 A. Gallo, G. Mani / Applied Su

.12. Statistical analysis

A one-way analysis of variance (ANOVA) was carried out and difference of p < 0.05 was considered statistically significant.he experimental data collected in this study are presented asean ± standard deviation.

. Results

.1. FTIR characterization

Fig. 3A shows the FTIR spectrum of the PAA coated stent (CS#2).he IR peaks observed between 1300 and 900 cm−1 provides valu-ble information regarding how the phosphonic acids bond withetal oxide surfaces. In the literature, the peak positions for the

R stretching vibrations of P OH, P O metal, and P O bonds haveeen observed between 900 and 950 cm−1, 1010 and 1050 cm−1,nd 1140 and 1250 cm−1, respectively [40–44]. In this study, theeaks observed at 932 cm−1 and 1148 cm−1 were assigned to P OHnd P O bonds of the phosphonate group of PAA deposited onhe stent, respectively. The strong peak observed at 1021 cm−1

as assigned to the P O metal bond, which suggests the cova-ent attachment of PAA on Co–Cr alloy stents. The deformationibration of the CH2 groups of PAA was observed at the peakosition 1368 cm−1. The C O stretching of the COOH group ofAA was observed at the peak position 1716 cm−1. In the litera-ure, the peak position of the C O stretching in COOH group haseen used to determine whether the COOH groups are free (notonded) or bonded with other functional groups [43]. The peakosition have been observed at ∼1710 cm−1 for the free COOHroups while the peak position has been shifted to a much loweravenumber ∼1680 cm−1 when the COOH groups were bondedith other functional groups [43]. Based on this, the C O peak posi-

ion (1716 cm−1) observed in this study suggests that the COOHroups of PAA deposited on the stent are free and available to formydrogen bonding with therapeutic drugs (PAT or DETA NONOate)sed in this study. The IR absorbance spectra obtained were nega-ive for all the stent samples. This could be due to the optical effectsf ultra-thin stent struts. Negative IR spectra for low reflectiveubstrates have been theoretically predicted and experimentallybserved in several studies [45–49]. Also, the polarization of an inci-ent light and the angle of incidence play a vital role in deciding theature (positive or negative) of IR spectra [45,46,48,49]. Althoughhe spectra obtained were negative, the IR peak positions observedor the PAA and the drug coatings were in excellent agreement withhe literature [40–44,50–52].

Fig. 3B shows the FTIR spectrum of PAT deposited on the ablu-inal surface of ABL-PAT. The fingerprint regions of the PAT were

bserved at the peak positions 671, 1073, and 1227 cm−1. Thetrong peak for the C O stretches of the ester groups in PAT wasbserved at 1712 cm−1. Also, the peak for the CH3 bending vibra-ions of PAT was observed at 1365 cm−1. These IR peak positionsere in agreement with the literature for the PAT coating [50–52].

ig. 3C shows the FTIR spectrum of DETA NONOate deposited onhe luminal surface of LUM-NO. The fingerprint regions of DETAONOate were observed at 669, 878, 938, and 1153 cm−1. Theeak for the scissoring vibration of CH2 groups was observedt 1460 cm−1. The peaks for the N O and N O stretches werebserved at 1550 and 1600 cm−1, respectively. A broad peak forhe NH3

+ stretching was observed at 2929 cm−1. Also, the sym-etric and asymmetric stretches of N H groups were observed at

250 cm−1 and 3309 cm−1, respectively. These results suggested

he coating of DETA NONOate on the luminal stent surfaces.

Fig. 3D and E shows the FTIR spectra of the abluminal and lumi-al surfaces of Co-PAT-NO, respectively. The IR peaks observedtrongly showed the presence of PAT and DETA NONOate on the

Science 279 (2013) 216– 232

abluminal and luminal surfaces of the stent, respectively. The IRpeak positions for the PAT on the abluminal surface and DETANONOate on the luminal surface were in excellent agreement withthose of PAT and DETA NONOate as provided in the above paragraphfor ABL-PAT (Fig. 3B) and LUM-NO (Fig. 3C) as well as with the lit-erature [50–52]. These results showed the successful co-coating ofPAT and DETA NONOate on the abluminal and luminal surfaces ofthe stent, respectively.

3.2. SEM characterization

Fig. 4 shows the SEM images of control stents (CS#1 and CS#2).The abluminal and luminal surfaces of the chemically cleaned con-trol stent (CS#1) are shown in Fig. 4A and B, respectively. The stentsurfaces appear smooth with no surface defects. The abluminaland luminal surfaces of the PAA coated stent (CS#2) are shown inFig. 4C and D, respectively. No significant difference in the surfacemorphology was observed between CS#1 and CS#2.

Fig. 5(A–D) shows the PAT coating on the abluminal surfacesof the stent (ABL-PAT). PAT formed two different types of mor-phologies including thin film-like structure (Fig. 5A–C) and needleshaped crystals (Fig. 5D) on the stents. Irrespective of the mor-phologies, the PAT was deposited on the abluminal surface withits boundary lying either on the sides of stent struts (Fig. 5A and B)or at the interface between the abluminal surface and the side ofthe stent strut (Fig. 5C and D). No PAT was observed on the lumi-nal surfaces of the stent. Fig. 5E and F shows the luminal surfaceof DETA NONOate coated stent (LUM-NO). Unlike PAT, no specificmorphology was observed for DETA NONOate coating. This sug-gested that the DETA NONOate was present as a molecular coatingwhich is difficult to image using a SEM because of the instrument’slimited resolution for imaging molecular coatings. The characteri-zation technique such as FTIR is commonly used to determine thepresence of molecular coatings. The FTIR characterization of DETANONOate coating on the luminal stent surfaces was provided in theprevious Section 3.1.

Fig. 6 shows the SEM images of the co-coated stents (Co-PAT-NO). The coating of PAT on the abluminal stent surfaces as thin film-like structure and needle shaped crystals are shown in Fig. 6A andB, respectively. The arrows provided in these images (Fig. 6A and B)show the boundary of PAT coating to confirm that the drug coatingdid not extend up to the luminal stent surface. Fig. 6C shows theDETA NONOate coated luminal surface of the co-coated stent. A lowmagnification (250×) image of the co-coated stent was provided inFig. 6D to show that the drug coating was uniformly distributedon the stent surface. In this image, a single arrow shows the PATcoating on the abluminal surface while the double arrow showsthe DETA NONOate coated luminal stent surface. Thus, the SEMimages confirmed that the morphologies and distribution of thedrug coating on the co-coated stent (Co-PAT-NO) were identical tothat of the single drug coated stents (ABL-PAT and LUM-NO).

3.2.1. SEM characterization of the expanded stentsFig. 7 shows the SEM images of the stents expanded using an

angioplasty balloon catheter. Both low (100×) and high magnifi-cation (500× and 1500×) SEM images were captured to study theintegrity of co-coating. Fig. 7A shows the low magnification imageof the expanded control stent while the Fig. 7B and C shows thehigh magnification images of the abluminal and luminal surfacesof the expanded control stent, respectively. Fig. 7D shows the lowmagnification image of the expanded co-coated stent. Fig. 7E and Fshows the high magnification images of the abluminal surfaces of

the co-coated stent. The arrows provided in these images show theboundary of PAT coating on the abluminal surface. Fig. 7G shows thehigh magnification image of the luminal surface of the co-coatedstent. No delamination or any cracking of the drug coatings was

A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232 221

F T, (C)s

ote

3

iot

ig. 3. FTIR spectra of (A) PAA coated stent (CS#2), (B) abluminal surface of ABL-PAurface of Co-PAT-NO.

bserved on the stent surfaces. Thus, these results demonstratedhat the integrity of the co-coating was maintained during the stentxpansion procedure.

.3. Optical profilometry characterization

Fig. 8A and B shows the 3D optical profilometry topographymages (63.7 �m × 47.7 �m) of the abluminal and luminal surfacesf CS#1, respectively. These images clearly showed the microstruc-ure of stent surfaces with a network of grains separated by grain

luminal surface of LUM-NO, (D) abluminal surface of Co-PAT-NO, and (E) luminal

boundaries. Table 4 shows the roughness (Sa) values measured byoptical profilometry for both the abluminal and luminal surfaces ofall the five different groups of stents used in this study. Sa is definedas the average roughness evaluated over the complete 3D surface.The Sa values of abluminal and luminal surfaces of the CS#1 weremeasured as 0.005 ± 0.001 �m and 0.006 ± 0.001 �m, respectively.

Fig. 8C and D shows the images of abluminal and luminal surfacesof CS#2, respectively. The surface topography of CS#2 was similarto that of CS#1. In the literature, when the molecular coatingswere deposited on the metal surfaces, the roughness values were

222 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

F ace ofC

muhnftrsipamtwooi

TAa

ig. 4. SEM images of (A) abluminal surface of control stent (CS#1), (B) luminal surfS#2.

easured before and after the deposition to determine the coatingniformity [40–42]. A significant difference in the roughness valueas been attributed to the formation of non-uniform coating whileo significant different in the roughness value was attributed the

ormation of uniform coating which followed the underlying metalopography [40–42]. In this study, no significant difference in theoughness values was observed between CS#1 and CS#2. Thisuggests that the PAA coating on the stent surface was uniform andt followed the contour of stent surfaces. Fig. 9 shows the opticalrofilometry images obtained after the deposition of drugs. Fig. 9And B shows the thin film-like morphology and needle shapedorphology of PAT on the abluminal surfaces of the stent, respec-

ively. In both the images, the underlying metal microstructureas not visible which suggested that the PAT was uniformly coated

n the abluminal stent surfaces. In the literature, the depositionf needle shaped PAT crystals on the metal surfaces significantlyncreased the surface roughness due to its rough morphology

able 4verage roughness values of stents measured by 3D optical surface profilometer for

scan size of 63.7 �m × 47.7 �m.

Stents Sa (�m)

Abluminal surface Luminal surface

CS#1 0.005 ± 0.001 0.006 ± 0.001CS#2 0.005 ± 0.000 0.005 ± 0.001ABL-PAT 0.074 ± 0.182 0.022 ± 0.019LUM-NO 0.004 ± 0.000 0.005 ± 0.001Co-PAT-NO 0.064 ± 0.080 0.013 ± 0.007

CS#1, (C) abluminal surface of PAA coated stent (CS#2), and (D) luminal surface of

[53]. In agreement with the literature, a significant increase in thesurface roughness value was observed for the abluminal surfaceof ABL-PAT when compared to that of the abluminal surfaces ofcontrol surfaces. The topography image of the luminal surfaceof ABL-PAT showed the microstructural grain features. Also, nosignificant increase in the surface roughness value was observedfor the luminal surface of ABL-PAT when compared to that of theluminal surfaces of control stents. These results strongly suggestedthat the PAT was not present on the luminal stent surface. Fig. 9Dand E shows the topography images of the abluminal and luminalsurfaces of the LUM-NO. In agreement with the SEM data, no signif-icant difference in the surface topography was observed betweenCS#2 and LUM-NO. This suggested that the DETA NONOate wasdeposited as a molecular coating which followed the contour ofmicrostructural grain features of the stent surfaces.

The topography images of the abluminal and luminal surfacesof the co-coated stent are shown in Fig. 10. The drug morpholo-gies (thin film-like PAT in Fig 10A; needle shaped PAT crystals inFig. 10B; DETA NONOate molecular coating in Fig. 10C) observed inco-coated stents were consistent with that of single drug coatedABL-PAT and LUM-NO stents. The luminal surfaces of LUM-NO(Fig. 9E) and Co-PAT-NO (Fig. 10C) look different because of thefollowing reason. For Co-PAT-NO, after PAT coating on the ablumi-nal surface, the luminal surface alone was cleaned using an ethanolwetted mandrel as described in section 2.4. However, no such pro-

cedure was carried out for LUM-NO since there was no PAT coatingon the abluminal stent surface. Hence, the luminal surface of Co-PAT-NO appears rougher (mainly because of the luminal cleaningprocedure employed) when compared to that of LUM-NO.

A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232 223

AT, an

3

aTC(dsthaai

Fig. 5. (A–D) SEM images of abluminal surfaces of ABL-P

.4. Contact angle characterization

Fig. 11 shows the images of contact angles obtained for thebluminal and luminal surfaces of the stents used in this study.he contact angles of the abluminal and luminal surfaces of theS#1 were measured as 104.1 ± 1.9◦ and 87 ± 5.5◦, respectivelyFig. 11A and B). After PAA coating, the contact angles significantlyecreased to 79.2 ± 3.7◦ and 76.1 ± 4◦ for the abluminal and luminaltent surfaces, respectively (Fig. 11C and D). A decrease in the con-act angle after PAA coating was expected since the PAA contains

ydrophilic COOH terminal groups (Fig. 1A). After PAT coating,n increase in the contact angle (95.2 ± 7.8◦) was observed for thebluminal surface of ABL-PAT (Fig. 11E). Although PAT is primar-ly a hydrophobic drug containing several aromatic rings and–CH3

d (E and F) SEM images of luminal surfaces of LUM-NO.

functional groups, there are few hydrophilic functional groups suchas OH, C O, COO, and NH are also present in its chemical struc-ture (Fig. 1B). Hence the contact angle of PAT can vary from 80◦ to100◦ depending on the orientation of different functional groupsand the type of morphology that the PAT crystals can form on amaterial surface. No significant difference in the contact angle wasobserved for the luminal surfaces of ABL-PAT (74.9 ± 3.6◦) and CS#2(76.1 ± 4◦), which suggested that the PAT was not present on theluminal stent surface. For LUM-NO, the luminal surface (60.6 ± 4.7◦)was more hydrophilic than that of CS#2 (76.1 ± 4◦). This is because

the DETA NONOate is primarily a hydrophilic drug containing sev-eral hydrophilic functional groups ( NH2, N O, NH3

+, and NO−)in its chemical structure (Fig. 1C). No significant difference in thecontact angle was observed between the abluminal surfaces of

224 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

F rface

s

LomJtDr

3

lsrrrfppoo3

TTs

ig. 6. SEM images of (A and B) abluminal surfaces of Co-PAT-NO, (C) luminal suurfaces of Co-PAT-NO.

UM-NO (82.3 ± 8.7◦) and CS#2 (79.2 ± 3.7◦). The contact anglesf the abluminal and luminal surfaces of the co-coated stents wereeasured as 82.9 ± 6.3◦ and 69.7 ± 11.2◦, respectively (Fig. 11I and

). In agreement with the other characterization techniques, con-act angle values also showed the successful deposition of PAT andETA NONOate on the abluminal and luminal surfaces of the stent,

espectively.

.5. Drug release studies

The total amounts of PAT and NO loaded on the abluminal anduminal stent surfaces, respectively, are provided in Table 5. Fig. 12Ahows the cumulative PAT release profile for ABL-PAT. A biphasicelease profile with an initial burst followed by a slow and sustainedelease was observed. Fig. 12B shows the actual amount of PATeleased between every two consecutive time points. In this figure,rom ‘Hour-1′ to ‘Hour-3 to Hour-6′ were plotted with respective torimary Y-axis while ‘Hour-6 to Hour-12′ to ‘Day-14 to Day-28′ were

lotted with respective to secondary Y-axis. An initial burst releasef 1.12 ± 0.3 �g on the first hour was followed by a sustained releasef 0.24 ± 0.1, 0.18 ± 0.1, 0.05 ± 0.01, and 0.03 ± 0.01 �g on hours –, 6, 12, and 24, respectively. After day-1, an amount closer to 30 ng

able 5otal amount of paclitaxel and nitric oxide loaded on the abluminal and luminaltent surfaces, respectively.

Total amount of paclitaxel loaded on theabluminal stent surface

2.2 ± 0.4 �g

Total amount of nitric oxide loaded on theluminal stent surface

2.9 ± 0.6 �M

of Co-PAT-NO, (D) low magnification image to show both abluminal and luminal

was sustained released between every two time points that wereused in the study for up to 14 days and a 80 ng of PAT was releasedbetween day-14 and day-28. Fig. 12C shows the cumulative nitricoxide (NO) release profile for the LUM-NO. All the NO coated wasburst released by the hour-1. Fig. 12D and E shows the PAT releaseprofile and the amount of PAT released between every two con-secutive time points for Co-PAT-NO, respectively. Fig. 12F showsthe NO release profile for Co-PAT-NO. Similar to the single drugcoated stents (ABL-PAT and LUM-NO), the PAT showed a biphasicdrug release profile with an initial burst on the first hour followedby a sustained release for up to 28 days while the NO was burstreleased by the first hour. Thus, the PAT and NO were co-deliveredfrom the abluminal and luminal surfaces of the stent, respectively.

4. Discussion

Late stent thrombosis (LST) is the limitation of currently avail-able drug-eluting stents [10–13]. LST is mainly caused due to thepoor or impaired endothelial cell growth on the luminal stentsurfaces [11–13]. Normal endothelial cells produce different anti-coagulant factors to inhibit adhesion, aggregation, and activation ofblood platelets and thereby preventing thrombosis [18]. However,when these cells are damaged or are not present on the luminalstent surfaces, it initiates the coagulation cascade which results instent thrombosis [11–13].

Most drug-eluting stents (DES) are coated with anti-

proliferative drugs on both the abluminal and luminal surfaces fortreating neointimal hyperplasia [10,14]. These anti-proliferativedrugs are not cell specific. Hence, when these drugs are released,they not only inhibit the growth of smooth muscle cells but also

A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232 225

Fig. 7. SEM images of (A–C) expanded control stents (CS#1), (D–G) expanded co-coated stents (Co-PAT-NO).

226 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

F inal

eohswbacgoes

cptfs[mdpHaTssad

ppP

ig. 8. Optical surface profilometry images of (A) abluminal surface of CS#1, (B) lum

ndothelial cells. Incomplete endothelial cell coverage observedn the luminal surfaces of drug-eluting stents in animals andumans has led to LST [11–13]. Hence, there is a need to developtrategies to promote endothelialization of luminal stent surfacehile inhibiting neointimal hyperplasia. This can be achieved

y co-delivering two different therapeutic agents, one from thebluminal stent surface to inhibit the growth of smooth muscleells, and the other from the luminal stent surface to promote therowth of endothelial cells. This study showed the co-deliveryf paclitaxel (an anti-proliferative agent) and nitric oxide (anndothelialization promoting agent) from abluminal and luminaltent surfaces, respectively.

Some polymeric drug delivery carriers used in stents haveaused adverse reactions [19–22]. Hence, different types ofolymer-free drug delivery approaches are currently under inves-igation for stents. These approaches can be classified under theollowing three different categories: (a) physical modification oftents such as porous [54], textured [55], and reservoir surfaces56]; (b) chemical modification of stents using self-assembled

onolayers (SAMs) [53,57,58]; (c) no modifications with therug directly coated onto stent material surfaces [36]. The phos-honoacetic acid (PAA) coating belongs to the second category.owever, the PAA coating is not a SAM since it does not have

hydrocarbon tail, which is a mandatory requirement for SAMs.he reasons for choosing PAA to form molecular coatings in thistudy are that the PAA contains phosphonic acid groups which cantrongly bind to different metal oxides leaving the COOH groupst the surface which can form hydrogen bonding with a variety ofrugs.

In this study, the FTIR showed the formation of metal-hosphonate bond between the stent and PAA as well as theresence of free COOH groups on stent surfaces. The nature ofAT coating on COOH terminated surfaces has been previously

surface of CS#1, (C) abluminal surface of CS#2, and (D) luminal surface of CS#2.

reported [53]. Briefly, the PAT contains both hydrogen bond donors( OH groups) and hydrogen bond acceptors (C O functionalities inester, ketone and amide groups). The COOH groups on the metalsurfaces also contain OH and C O functionalities. When the PAThas been allowed to deposit on COOH terminated metal surfaces,the OH functionalities (of COOH groups) on metal surfaces canform hydrogen bonds with–OH groups of PAT, NH groups of PAT,and C O functionalities in the ester, ketone, and amide groups ofPAT. Similarly, the C O functionalities (of COOH groups) on metalsurfaces can form hydrogen bonds with OH groups of PAT. Thus,the PAT can first form a molecular coating directly on the metal sur-faces due to the hydrogen bonding interactions between the PATmolecules and the COOH groups on the metal surface [53]. Next,an extensive intermolecular hydrogen bonding occurs between thePAT molecules to form crystals deposited on top of the PAT molec-ular coating [36,53]. The same mechanism is proposed here forthe deposition of PAT on the PAA coated stent surfaces. PAT formsdifferent types of morphologies depending on the nature of sol-vent used to dissolve the drug, the method of deposition, and thematerial surface on which the drug is deposited [36,50,52,53]. Ithas formed needle [36,53], spherical [36], ovoid [50], particle [50],and powder-shaped [53] morphologies on flat Co–Cr alloy surfaceswhen the drug was microdeposited using ethanol (100%) as a sol-vent. On microrough Co–Cr alloy surfaces, the drug did not formany specific morphology as it just filled the cavities of the roughsurfaces [51]. In this study, the thin film- and needle-shaped PATmorphologies were observed on the stent surfaces when the drugwas spray coated using the solvent composition of 75% ethanol and25% DMSO.

The release of PAT has been achieved due to the breakage ofhydrogen bonds between the PAT molecules and metal surfaces,and between the PAT molecules in the crystals by the ions in physi-ological solution such as PBS/T-20 [36,53]. A biphasic release profile

A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232 227

Fig. 9. Optical surface profilometry images of (A and B) abluminal surfaces of ABL-PAT, (C) luminal surface of ABL-PAT, (D) abluminal surface of LUM-NO, and (E) luminalsurface of LUM-NO.

228 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

nal su

hmbwsbo

2Hwmcwu(puhrcdtppsTdBr

Fig. 10. Optical surface profilometry images of (A,B) ablumi

as been observed for the delivery of PAT from the COOH ter-inated metal surfaces [53]. The PAT crystals are usually weakly

ound to each other and are burst released while the PAT moleculeshich are hydrogen bonded to metal surfaces are released at a

low and sustained rate [51,53]. In agreement with these studies, aiphasic release profile was observed in this study for the deliveryf PAT from the PAA coated abluminal stent surfaces.

In this study, the total amount of PAT coated on the stents was.2 ± 0.4 �g. The total surface area of the stents is 64.9352 mm2.ence, the drug dosage is 0.034 ± 0.006 �g/mm2. This is lesserhen compared to the dosage of 1 �g/mm2 loaded on the com-ercially available drug-eluting stents. However, since the drug is

oated only on the abluminal surface, it is expected that the dosageill be lesser when compared to the ones that are commerciallysed. Serruys et al. [59] have evaluated the effect of different doses10 or 30 �g/stent) of PAT on neointimal hyperplasia in humanatients. Although the doses used were only 10–30% of the dosessed in commercially available stents, the inhibition of neointimalyperplasia was comparable [59]. Also, the low dose (10 �g/stent)eleased over a period of 30 days was found to be very effective inontrolling the neointimal hyperplasia. Jabara et al. [60] have alsoemonstrated that a low dose of PAT (0.15 �g/mm2) suppressedhe neointima formation while decreasing the drug toxicity andromoting vessel healing. In another study, Shinke et al. [61] com-ared the efficacy of a very low dose (0.025 �g/mm2) of PAT coatedtent with a TAXUS Express stent (1 �g/mm2) in a porcine model.

he inhibition of neointima formation observed for the low PATose containing stent was equivalent to that of the TAXUS stent.ased on these literature, it is expected that the amount of PATeleased in this study would inhibit neointimal hyperplasia.

rfaces of Co-PAT-NO, and (C) luminal surface of Co-PAT-NO.

Similar to PAT, DETA NONOate can also form hydrogen bondingwith PAA coated luminal stent surfaces. This can occur betweenthe OH functionalities (of COOH groups) of PAA coated stentsurfaces and the NH, N O, and N O groups of DETA NONOate.Also, since the DETA NONOate contains both positively (NH3

+) andnegatively (NO−) charged functional groups, it can also form inter-molecular electrostatic interactions between the drug molecules.Hence, the hydrogen bonding and the electrostatic interactions areconsidered to be the bonds primarily responsible for the coatingof DETA NONOate on the stent surfaces. The mechanism proposedfor the delivery of DETA NONOate from the stent surfaces is thebreakage of hydrogen bonds between the DETA NONOate and thePAA coated stent surfaces, and the breakage of electrostatic attrac-tions between the drug molecules, by the ions in PBS/T-20 solution.Since DETA NONOate is a hydrophilic drug, the H2O molecules haveeasy access into the drug coating and participate in the immediatebreakage of hydrogen bonds and electrostatic interactions, whichcould be the reason for the burst release of this drug.

The release of nitric oxide from different nitric oxide donordrugs can occur either spontaneously or through biotransformation[62–65]. Diazeniumdiolates (NONOates) belongs to the first cate-gory which releases nitric oxide spontaneously under physiologicalconditions [62,65]. The DETA NONOate decomposes spontaneouslyat a physiological pH (7.4) and temperature (37 ◦C) into 2 molesof nitric oxide per mole of the parent compound [62]. The NOreleased from NONOates or other nitric oxide donor drugs have

significantly increased the growth of endothelial cells on differentmaterial surfaces [26,66]. The release of NO not only encouragesendothelialization [26,66] but also inhibits neointimal hyperplasia[29,66,67] and thrombosis [31,66]. Zhang et al. [68] have prepared

A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232 229

Fig. 11. Contact angles of the abluminal and luminal surfaces of CS#1, CS#2, ABL-PAT, LUM-NO, and Co-PAT-NO.

230 A. Gallo, G. Mani / Applied Surface Science 279 (2013) 216– 232

Fig. 12. Drug release profiles of (A and B) ABL-PAT, (C) LUM-NO, (D–F) Co-PAT-NO.

rface S

Nioocbapipse(gt(el

5

sapoagdoafwasil

A

g

R

[

[

[

[[

[

[

[[[

[

[

[

[

[

[

[

[[

[[

[

[

[[

[

[

[[[[[[

[

[

[[

[[[[[[[[

[

[

[

A. Gallo, G. Mani / Applied Su

O releasing silica particles using a NONOate coating. Under phys-ological conditions (in PBS buffer, pH 7.4 at 37 ◦C), a burst releasef NO was observed from these particles and the total amountf NO coated (∼1.7 �M) was released by 10 h. The polymer filmsoated with these NO releasing particles showed excellent throm-oresistance properties when compared to the controls. Similarly,

total of 0.25–0.45 �M of NO released within a day successfullyrovided thromboresistance properties to the polymers for use

n catheter-style oxygen sensors [69]. Taite et al. [26] have pre-ared novel polyurethanes containing NO donor functional groupsuch as NONOates and investigated the effect of NO release on thendothelialization of the polymer. Although 70% of the NO loaded∼0.15 �M) was burst released from the polymer, these surfacesreatly favored the growth of endothelial cells when compared tohe controls. Based on these literature, it is expected that the NO∼3.0 �M) released in this study would encourage the growth ofndothelial cells and provide anti-thrombogenic properties to theuminal stent surfaces.

. Conclusions

A bidirectional dual drug-eluting stent was prepared in thistudy by co-coating paclitaxel and nitric oxide donor drug on thebluminal and luminal stent surfaces, respectively. A polymer-freelatform using phosphonoacetic acid was used to co-coat the drugsnto stent surfaces. The surface characterization techniques suchs FTIR, SEM, 3D optical surface profilometry, and contact angleoniometry confirmed the co-coating of drugs on stent surfaces. Noelamination or cracks formation was observed on the co-coatingf drugs during the stent expansion. PAT was released from thebluminal stent surfaces in a biphasic manner with an initial burstollowed by a slow and sustained release while the nitric oxideas burst released from the luminal stent surfaces. Thus, paclitaxel

nd nitric oxide were co-delivered from the abluminal and luminaltent surfaces, respectively. This stent has potential applications fornhibiting neointimal hyperplasia and promoting luminal endothe-ialization to prevent late stent thrombosis.

cknowledgement

This study was supported by a competitive research grant pro-ram from the South Dakota Board of Regents (SDBOR).

eferences

[1] P.W. Wilson, R.B. D’Agostino, D. Levy, A.M. Belanger, H. Silbershatz, W.B. Kannel,Circulation 97 (1998) 1837–1847.

[2] R.F.J. Shepherd, R.E. Vlietstra, The history of balloon angioplasty, in: R.E. Vliet-stra, D.R. Holmes (Eds.), Percutaneous Transluminal Coronary Angioplasty, F.A.Davis Company, Philadelphia, 1987, pp. 1–17.

[3] E.J. Topol, Textbook of Interventional Cardiology, W.B. Saunders, Philadelphia,1994.

[4] P.W. Serruys, P.D. Jaegere, F. Kiemeneij, C. Macaya, W. Rutsch, G. Heyndrickx, H.Emanuelsson, J. Marco, V. Legrand, P. Materne, J. Belardi, U. Sigwart, A. Colombo,J.J. Goy, P.V.D. Heuvel, J. Delcan, M.A. Morel, New England Journal of Medicine331 (1994) 489–495.

[5] D.L. Fischman, M.B. Leon, D.S. Baim, R.A. Schatz, M.P. Savage, I. Penn, K. Detre, L.Veltri, D. Ricci, M. Nobuyoshi, M. Cleman, R. Heuser, D. Almond, P.S. Teirstein,R.D. Fish, A. Colombo, J. Brinker, J. Moses, A. Shaknovich, J. Hirshfeld, S. Bai-ley, S. Ellis, R. Rake, S. Goldberg, New England Journal of Medicine 331 (1994)496–501.

[6] N. Kipshidze, G. Dangas, M. Tsapenko, J. Moses, M.B. Leon, M. Kutryk, P. Serruys,Journal of the American College of Cardiology 44 (2004) 733–739.

[7] A.C. Newby, A.B. Zaltsman, Journal of Pathology 190 (2000) 300–309.[8] N. Kipshidze, M. Leon, M. Tsapenko, R. Falotico, G. Kopia, J. Moses, Current

Pharmaceutical Design 10 (2004) 337–348.

[9] S. Tanimoto, J. Daemen, P.W. Serruys, Journal of Vascular Health and Risk Man-

agement 3 (2007) 481–490.10] S. Garg, P.W. Serruys, Journal of the American College of Cardiology 56 (2010)

S1–S42.11] S. Windecker, B. Meier, Circulation 116 (2007) 1952–1965.

[

[

cience 279 (2013) 216– 232 231

12] A.V. Finn, M. Joner, G. Nakazawa, F. Kolodgie, J. Newell, M.C. John, H.K. Gold, R.Virmani, Circulation 115 (2007) 2435–2441.

13] T. Pilgrim, S. Windecker, Minerva Cardioangiologica 57 (2009) 611–620.14] S. Garg, P.W. Serruys, Journal of the American College of Cardiology 56 (2010)

S43–S78.15] M. Joner, G. Nakazawa, A.V. Finn, S.C. Quee, L. Coleman, E. Acampado, P.S. Wil-

son, K. Skorija, Q. Cheng, X. Xu, H.K. Gold, F.D. Kolodgie, R. Virmani, Journal ofthe American College of Cardiology 52 (2008) 333–342.

16] M. Joner, A. Finn, A. Farb, E. Mont, F. Kolodgie, E. Ladich, R. Kutys, K. Skorija,H. Gold, R. Virmani, Journal of the American College of Cardiology 48 (2006)203–205.

17] G. Nakazawa, A.V. Finn, R. Virmani, HERZ 32 (2007) 274–280.18] K.K. Wu, P. Thiagarajan, Annual Review of Medicine 47 (1996) 315–331.19] R. Virmani, G. Guagliumi, A. Farb, G. Musumeci, N. Grieco, T. Motta, L. Mihalcsik,

M. Tespili, O. Valsecchi, F. Kolodgie, Circulation 109 (2004) 701–705.20] R. Virmani, A. Farb, G. Guagliumi, F. Kolodgie, Coronary Artery Disease 15 (2004)

313–318.21] R. Virmani, F. Kolodgie, A. Farb, The American Heart Hospital Journal 2 (2004)

85–88.22] J.R. Nebeker, R. Virmani, C.L. Bennet, J.M. Hoffman, M.H. Samore, J. Alvarez, C.J.

Davidson, J.M. Mckoy, D.W. Raish, B.K. Whisenant, P.R. Yarnold, S.M. Belknap,D.P. West, J.E. Gage, R.E. Morse, G. Gligoric, L. Davidson, M.D. Feldman, Journalof the American College of Cardiology 47 (2006) 182–183.

23] D.I. Axel, W. Kunert, C. Göggelmann, M. Oberhoff, C. Herdeg, A. Küttner, D.H.Wild, B.R. Brehm, R. Riessen, G. Köveker, K.R. Karsch, Circulation 96 (1997)636–645.

24] A. Heldman, L. Cheng, G. Jenkins, P. Heller, D. Kim, M. Ware, C. Nater, R. Hruban,B. Rezai, B. Abella, K. Bunge, J. Kinsella, S. Sollott, E. Lakatta, J. Brinker, W. Hunter,J. Froehlich, Circulation 103 (2001) 2289–2295.

25] J. Waugh, A.J. Wagstaff, American Journal of Cardiovascular Drugs 4 (2004)257–268.

26] L.J. Taite, P. Yang, H.-W. Jun, J.L. West, Journal of Biomedical Materials Research-Applied Biomaterials 84 (2008) 108–116.

27] H.W. Jun, L.J. Taite, J.L. West, Biomacromolecules 6 (2005) 838–844.28] L. Taite, J.L. West, Cardiovascular Engineering and Technology 2 (2011)

113–123.29] E.A. Lipke, J.L. West, Acta Biomaterials 1 (2005) 597–606.30] D. Hou, H. Narciso, K. Kamdar, P. Zhang, B. Barclay, K.L. March, Cardiovascular

and Interventional Radiology 28 (2005) 60–65.31] K.A. Mowery, M.H. Schoenfisch, J.E. Saavedra, L.K. Keefer, M.E. Meyerhoff, Bio-

materials 21 (2000) 9–21.32] L. Sipos, A. Som, R. Faust, R. Richard, M. Schwarz, S. Ranade, M. Boden, K. Chan,

Biomacromolecules 6 (2005) 2570–2582.33] T.G. Kim, H. Lee, Y. Jang, T.G. Park, Biomacromolecules 10 (2009) 1532–1539.34] H.J. Lim, H.Y. Nam, B.H. Lee, D.J. Kim, J.Y. Ko, J.S. Park, Biotechnology Progress

23 (2007) 693–697.35] F. Innocente, D. Mandracchia, E. Pektok, B. Nottelet, J.C. Tille, S.D. Valence, G.

Faggian, A. Mazzucco, A. Kalangos, R. Gurny, M. Moeller, B.H. Walpoth, Circu-lation 120 (2009) S37–S45.

36] G. Mani, C.E. Macias, M.D. Feldman, D. Marton, S. Oh, C.M. Agrawal, Biomaterials31 (2010) 5372–5384.

37] Z.H. Taha, Talanta 61 (2003) 3–10.38] N.S. Kishnani, H.L. Fung, Methods in Enzymology 268 (1996) 259–265.39] J. Dortch-Carnes, K. Russell, Experimental Eye Research 84 (2007) 185–190.40] A. Raman, M. Dubey, I. Gouzman, E.S. Gawalt, Langmuir 22 (2006) 6469–6472.41] R. Quinones, E.S. Gawalt, Langmuir 23 (2007) 10123–10130.42] A. Raman, R. Quinones, L. Barriger, R. Eastman, A. Parsi, E.S. Gawalt, Langmuir

26 (2010) 1747–1754.43] S. Mohapatra, P. Pramanik, Colloids and Surfaces A: Physicochemical and Engi-

neering Aspects 339 (2009) 35–42.44] V. Zima, J. Svoboda, L. Benes, K. Melanova, M. Trchova, A. Ruzicka, Journal of

Solid State Chemistry 182 (2009) 3155–3161.45] A. Udagawa, T. Matsui, S. Tanaka, Applied Spectroscopy 40 (1986) 794–799.46] A. Gericke, A.V. Michailov, H. Huhnerfuss, Vibrational Spectroscopy 4 (1993)

335–348.47] Y.S. Yen, J.S. Wong, Journal of Physical Chemistry 93 (1989) 7208–7216.48] J.A. Mielczarski, R.H. Yoon, Journal of Physical Chemistry 93 (1989) 2034–2038.49] J.A. Mielczarski, Journal of Physical Chemistry 97 (1993) 2649–2663.50] S.E. Stoebner, G. Mani, Applied Surface Science 258 (2012) 5061–5072.51] S. Lancaster, S. Kakade, G. Mani, Langmuir 28 (2012) 11511–11526.52] A.B. Dhanikula, R. Panchagnula, The AAPS Journal 6 (2004) 1–12.53] G. Mani, N. Torres, S. Oh, Biointerphases 6 (2011) 33–42.54] I. Tsujino, J. Ako, Y. Honda, P.J. Fitzgerald, Expert Opinion on Drug Delivery 4

(2007) 287–295.55] R. Wessely, J. Hausleiter, C. Michaelis, B. Jaschke, M. Vogeser, S. Milz, B.

Behnisch, T. Schratzenstaller, M.R. Gluszko, M. Stöver, E. Wintermantel, A. Kas-trati, A. Schömig, Arteriosclerosis, Thrombosis, and Vascular Biology 25 (2005)748–753.

56] A. Abizaid, J.R. Costa, F. Feres, Catheterization and Cardiovascular Interventions77 (2011) 49–51.

57] G. Mani, D.M. Johnson, D. Marton, M.D. Feldman, D. Patel, A. Ayon, C.M. Agrawal,

Biomaterials 29 (2008) 4561–4573.

58] G. Mani, B. Chandrasekar, M.D. Feldman, D. Patel, C.M. Agrawal, Journal ofBiomedical Materials Research B-Applied Biomaterials 90 (2009) 789–801.

59] P.W. Serruys, G. Sianos, A. Abizaid, J. Aoki, P.D. Heijer, H. Bonnier, P. Smits, D.McClean, S. Verheye, J. Belardi, J. Condado, M. Pieper, L. Gambone, M. Bressers,

2 rface

[

[

[[

[

[

[[

[68] H. Zhang, G.M. Annich, J. Miskulin, K. Stankiewicz, K. Osterholzer, S.I. Merz, R.H.

32 A. Gallo, G. Mani / Applied Su

J. Symons, E. Sousa, F. Litvack, Journal of the American College of Cardiology 46(2005) 253–260.

60] R. Jabara, N. Chronos, D. Conway, W. Molema, K. Robinson, JACC: CardiovascularInterventions 1 (2008) 81–87.

61] T. Shinke, S. Geva, L. Pendyala, R. Jabara, J. Li, J.P. Chen, A. Venegoni, K. Colley, R.Klein, N. Chronos, K. Robinson, D. Hou, International Journal of Cardiology 135(2009) 93–101.

62] M.R. Miller, I.L. Megson, British Journal of Pharmacology 151 (2007) 305–321.63] G.R. Thatcher, A.C. Nicolescu, B.M. Bennett, V. Toader, Free Radical Biology and

Medicine 37 (2004) 1122–1143.64] D. Schade, J. Kotthaus, B. Clement, Pharmacology and Therapeutics 126 (2010)

279–300.

[

Science 279 (2013) 216– 232

65] L.K. Keefer, R.W. Nims, K.W. Davies, D.A. Wink, Methods in Enzymology 268(1996) 281–293.

66] M.C. Frost, M.M. Reynolds, M.E. Meyerhoff, Biomaterials 26 (2005) 1685–1693.67] Y.S. Do, E.Y. Kao, F. Ganaha, H. Minamiguchi, K. Sugimoto, J. Lee, C.J. Elkins, P.G.

Amabile, M.D. Kuo, D.S. Wang, J.M. Waugh, M.D. Dake, Radiology 230 (2004)377–382.

Bartlett, M.E. Meyerhoff, Journal of the American Chemical Society 125 (2003)5015–5024.

69] M.H. Schoenfisch, K.A. Mowery, M.V. Rader, N. Baliga, J.A. Wahr, M.E. Meyerhoff,Analytical Chemistry 72 (2000) 1119–11126.