J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 5 , N O . 6 , 2 0 1 2

© 2 0 1 2 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 9 8 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c i n . 2 0 1 1 . 1 2 . 0 1 7

Pathology of Drug-Eluting Versus Bare-Metal Stentsin Saphenous Vein Bypass Graft Lesions

Saami K. Yazdani, PHD,* Andrew Farb, MD,† Masataka Nakano, MD,* Marc Vorpahl, MD,*Elena Ladich, MD,* Aloke V. Finn, MD,‡ Frank D. Kolodgie, PHD,* Renu Virmani, MD*

Gaithersburg and Silver Spring, Maryland; and Atlanta, Georgia

Objectives The purpose of this study was to assess the pathological responses of atheroscleroticsaphenous vein bypass grafts (SVBGs) to drug-eluting stents (DES) versus bare-metal stents (BMS).

Background Repeat bypass surgery is typically associated with a high rate of morbidity and mortal-ity. Percutaneous coronary interventions have emerged as the preferred treatment; however, onlylimited data are available on SVBGs pathological responses to DES and BMS.

Methods Formalin-fixed SVBG of �2 years duration (n � 31) were collected to histologically charac-terize advanced atherosclerotic lesions in native SVBG. In a separate group, SVBGs treated with DES(n � 9) and BMS (n � 9) for �30 days duration were assessed for morphological and morphometricchanges.

Results Necrotic core lesions were identified in 25% of SVBG sections, and plaque rupture with lu-minal thrombosis was observed in 6.3% of histological sections (32% [10 of 31] vein grafts exam-ined). Morphometry of DES demonstrated reduction in neointimal thickening versus BMS (0.13 mm[interquartile range: 0.06 to 0.16 mm] vs. 0.30 mm [interquartile range: 0.20 to 0.48 mm], p � 0.004).DES lesions also showed greater delayed healing characterized by increased peristrut fibrin deposi-tion, higher percentage of uncovered struts, and less endothelialization compared with BMS. Stentfractures (DES 56% vs. BMS 11%, p � 0.045) and late stent thrombosis (DES 44% vs. BMS 0%,p � 0.023) were more common in DES versus BMS.

Conclusions Advanced SVBG atherosclerotic lesions are characterized by large hemorrhagic necroticcores. Stenting of such lesions is associated with delayed vascular healing and late thrombosis par-ticularly following DES implantation, which may help explain the higher rates of cardiovascularevents observed in SVBG stenting as compared with native coronary arteries. (J Am Coll CardiolIntv 2012;5:666–74) © 2012 by the American College of Cardiology Foundation

From the *CVPath Institute Inc., Gaithersburg, Maryland; †Office of Device Evaluation, Center for Devices and RadiologicalHealth, U.S. Food and Drug Administration, Silver Spring, Maryland; and the ‡Department of Cardiology, Emory UniversitySchool of Medicine, Atlanta, Georgia. Dr. Finn is supported by National Institutes of Health grant HL096970-01A1, the CarlyleFraser Heart Center at Emory University, sponsored research agreement with Medtronic and St Jude Medical, and is a consultantfor Abbott Vascular and Cordis. Dr. Virmani is a consultant for Medtronic AVE, Abbott Vascular, W.L. Gore, Atrium Medical,Arsenal Medical, and Lutonix. All other authors have reported that they have no relationships relevant to the contents of this paperto disclose.

Manuscript received August 10, 2011; revised manuscript received December 7, 2011, accepted December 22, 2011.

va

DwHtnew

ppSdD

M

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2 Yazdani et al.

J U N E 2 0 1 2 : 6 6 6 – 7 4 Stenting of Saphenous Vein Grafts

667

Saphenous vein bypass grafts (SVBGs) remain the mostfrequently used conduits for coronary artery bypass graft(CABG) surgery. These grafts are susceptible to the rapiddevelopment of atherosclerotic lesions, and repeat revascu-larization procedures are common after 5 years (1,2). Sincerepeat CABG surgery is associated with increased morbidityand mortality compared with the index bypass procedure,percutaneous coronary intervention has emerged as thepreferred treatment for atherosclerotic SVBG failure (3–6).Bare-metal stent (BMS) implantation in SVBG was shownto reduce clinical events as compared with balloon angio-plasty (7,8). However, results of BMS implantation inSVBGs are less favorable than those in native coronaryarteries, with restenosis rates exceeding 30%, and are asso-ciated with a high rate of major cardiovascular events (9).

Reduced restenosis rates associated with sirolimus (SES)(Cypher, Cordis, Johnson & Johnson, Miami, Florida) andpaclitaxel (PES) (Taxus, Boston Scientific Corporation, Boston,Massachusetts) drug-eluting stents in native coronary arterieshas resulted in the frequent use of drug-eluting stents(DES) in atherosclerotic SVBGs (10,11). Within 1 yearpost-procedure, DES implantation in SVBGs result in areduction in target lesion revascularization rates comparedwith BMS (DES: 4% vs. BMS: 11%, p � 0.22) (12);however, the long-term effectiveness of this benefit of DESversus BMS has been inconsistent (13). In a randomizedcontrolled trial of SVBG revascularization, DES had alower incidence of target vessel revascularization at 6months (DES: 4.3% vs. BMS: 24.5%, p � 0.005), but targetessel revascularization rates were not significantly differentt �2 years (DES: 36% vs. BMS: 41%, p � 0.64) (14–16).

Preclinical and human autopsy studies have shown thatES implanted in native coronary arteries are associatedith delayed arterial healing compared with BMS (17).owever, pathological studies of atherosclerotic SVBGs

reated with DES versus BMS are lacking, and it isoteworthy that the atherosclerotic process is highly accel-rated and has some morphological differences comparedith native coronary disease.The objectives of this study were to perform a detailed

athological assessment of advanced SVBG atheroscleroticlaques and to determine the vascular healing profile ofVBGs implanted with SES or PES of �30 days implanturation to gain insights into the safety and effectiveness ofES implantation.

ethods

SVBG atherosclerotic plaque assessment. Formalin-fixedhuman hearts from patients treated with CABG surgery atleast 2 years before death were selected from our CVPathdatabase. SVBGs were excised in their entirety, serially

sectioned transversely at 3-mm intervals, and then embed-

ded in paraffin. Each 5-�m section was stained withhematoxylin and eosin and Movat pentachrome stains.

The SVBG cross sections were categorized as nonathero-sclerotic (i.e., only fibrointimal hyperplasia present [fibro-intimal proliferation (FIT)]) or atherosclerotic. Atheroscle-rotic SVBG lesions were further categorized as: 1) intimalxanthoma or fatty streak lesions (foam cell accumulationlacking a fibrous cap or lipid core); 2) fibroatheroma (FA)lesions (plaques with well-defined necrotic cores with orwithout focal accumulations of foam cells); 3) hemorrhagicnecrotic core (HNC) lesions with intact fibrous caps; and4) plaque rupture with a disrupted cap and an overlyingthrombus. Morphometric analysis was used to measure thearea within the internal elastic lamina, lumen, total plaquearea, and total necrotic core area and to calculate areastenosis and the percentage of the plaque occupied by anecrotic core (18,19).Stented SVBGs. From our regis-try of �600 human coronaryartery stents submitted for con-sultation, SVBGs stented withDES implanted �30 days beforedeath were identified. Clinicalhistories and cardiac catheteriza-tion reports were reviewed whenavailable. The control groupconsisted of SVBGs treated withBMS with matched stentlengths and implant durations.

Stented segments of SVBGswere fixed in 10% buffered for-malin, dissected off the heart,radiographed, and processed forplastic embedding. Stented veingrafts were sawed consecutivelyfrom the proximal to distal endsof the stents at 2- to 3-mmintervals. The sections were cut at 6 �m and stained withhematoxylin and eosin and Movat pentachrome stains.

Acute luminal thrombus was defined as a platelet-richthrombus that occupied �30% of the cross-sectional area ofthe lumen, and restenosis was defined as �75% in-stentcross-sectional area luminal narrowing by neointima. Pre-mortem luminal thrombus was distinguished from post-mortem thrombus based on histological appearance aspre-mortem thrombus is composed of platelet aggregationwithout white cells, and a propagated thrombus is composedof an interspersed layer of fibrin and red cells, whereas apost-mortem thrombus lacks the layered appearance. Latestent thrombosis, consistent with the clinical definition, wasdefined as thrombosis within a stent implanted �30 daysbefore death (20). The cause of death was determined aftercomplete autopsy and was categorized as stent-related,

Abbreviationsand Acronyms

BMS � bare-metal stent(s)

CABG � coronary arterybypass graft

DES � drug-eluting stent(s)

FA � fibroatheroma(s)

FIT � fibrointimalproliferation

HNC � hemorrhagic necroticcore

IQR � interquartile range

PES � paclitaxel-elutingstent(s)

SES � sirolimus-elutingstent(s)

SVBG � saphenous veinbypass graft

non–stent-related cardiac, and non

cardiac (21).

taeitudcWs

R

htT98

3

fipadb

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2

J U N E 2 0 1 2 : 6 6 6 – 7 4

Yazdani et al.

Stenting of Saphenous Vein Grafts

668

Computer-guided morphometric measurements and his-tological characterization of the stented segments, includingfibrin, inflammation, strut coverage, strut malapposition,necrotic core strut penetration, neoatherosclerosis, has beenpreviously defined (17,21–23).Statistical analysis. Continuous variables with normal dis-ribution were expressed as mean � SD. Discrete variablesnd continuous variables with non-normal distribution werexpressed as median (interquartile range [IQR]). Compar-sons of continuous variables with normal distribution wereested by analysis of variance. A Wilcoxon rank sum test wassed to compare non-normally distributed parameters oriscrete variables, and the Pearson chi-square test was used forategorical variables. Normality of distribution was tested with the

ilk-Shapiro test. A value of p � 0.05 was considered statisticallyignificant.

esults

SVBG atherosclerotic plaque assessment. A total of 589istological cross sections from 31 SVBGs (nonstent-reated) from 16 patients were prepared and examined.here were 7 cardiac deaths (4 sudden cardiac deaths) andnoncardiac deaths. CABG surgery had been performed

.5 � 5.9 years before death (range: 2 to 22 years). Grossly,older vein grafts (�5 years) showed atherosclerotic change(yellow, with cheesy material and hemorrhage), whereasthose implanted for 2 to 5 years usually showed intimalthickening with or without foam cell infiltration and ab-sence of necrotic core. As shown in Table 1, incidence offibrointimal thickening was observed in most lesions(90.3%). Intimal xanthoma and FA with necrotic core wereobserved in 74.2% and 64.5% of the 31 SVBGs, respec-tively. Seventeen of the 31 lesions (54.8%) demonstratedHNCs, with only 10 lesions (32.3%) showing plaquerupture.

To further characterize these histological changes, histo-logical cross sections from all 31 SVBGs (n � 589 sections)were categorized. The results demonstrated that most sec-tions (56.5%) showed FIT, with the remaining cross sec-tions (43.5%) demonstrating atherosclerotic changes. Inti-mal xanthoma was identified in 18.8% of the total SVBG

Table 1. Lesion Morphology, Stenosis, and Lipid Core Size of SVBG (>2 Ye

Changes in SVBGIncidence per

Lesion (%)Number of

Sections (%)

FIT 28 (90.3) 333 (56.5) 10

IX 23 (74.2) 111 (18.8) 12

FA with necrotic core 20 (64.5) 70 (11.9) 14

HNC 17 (54.8) 44 (7.5) 1

PR 10 (32.3) 31 (6.3) 1

Values are n (%) or mean � SD.

FA � fibroatheroma; FIT � fibrointimal thickening; HNC � hemorrhagic necrotic core; IEL � internal ela

sections, followed by lesser incidences of necrotic core(11.9%), HNCs (7.5%), and plaque rupture and luminalthrombus (6.3%) (Table 1, Fig. 1). Ruptured plaque wasprimarily composed of a disrupted fibrous cap with lipid-laden macrophages close to underlying necrotic core andwithin the fibrous cap. The necrotic core was usually largewith numerous cholesterol clefts and hemorrhage, with theoccasional presence of focal calcification, usually close to thevein wall (Online Fig. 1). The cap was more discontinuousthan what is seen in native coronary artery disease. Thenecrotic core always showed a base of neointimal layercomposed of smooth muscle cells and collagenous matrix.The fibrous caps of plaque ruptures were significantlythinner compared with SVBGs with plaque hemorrhage(82 � 42 �m vs. 214 � 162 �m, respectively, p � 0.002).The mean longitudinal length of plaque ruptures (estimatedby the number of contiguous SVBG cross sectionsdemonstrating rupture) was 6.6 � 7.2 mm (range 3.0 to

0 mm).At the aortic anastomosis, 67% of sites showed only

brointimal proliferation, and 33% had atheroscleroticlaque (with or without lipid core formation). At the distalnastomosis of SVBG to the native coronary artery, only 4%emonstrated atherosclerotic plaque, and 96% showed fi-rointimal thickening.

Stented SVBGs. Eighteen stented SVBGs, including 9 DESfrom 7 patients and 9 BMS from 8 patients, were analyzed,with a mean patient age of 71 � 10 years and 67 � 6 years,respectively. The DES stents consisted of 6 PES and 3 SES.Implanted BMS were 5 Multi-Link (Abbott Vascular,Santa Clara, California), 2 AVE (Medtronic, Minneapolis,Minnesota), 1 Bx Velocity (Cordis Corp, Johnson & John-son, Warren, New Jersey), and 1 Palmaz-Schatz (Johnson& Johnson). The duration of stent implantation and stentlength were greater in DES versus BMS, but differences didnot reach statistical significance (Table 2).

Post-mortem radiographs demonstrated fractures in 5 (3SES and 2 PES) of the 9 DES and in 1 of 9 BMS (BxVelocity). Stent fractures ranged from grade II (multiplebreaks with preserved alignment) to grade IV (multiplebreaks with transection but without gap) (24). Most of the

ea)

Stenosis(%)

Lipid Core Area(mm2) % Lipid Core

2.28 34 � 15 — —

3.02 37 � 9 — —

2.59 46 � 17 1.5 � 1.9 18 � 19

2.72 52 � 19 2.8 � 3.3 28 � 31

3.56 75 � 24 9.3 � 7.9 57 � 51

ars)

IEL Ar(mm2

.97 �

.36 �

.08 �

4.62 �

5.35 �

stic lamina; IX � intimal xanthoma; PR � plaque rupture; SVBG � saphenous vein bypass graft.

s(

SfonbDowacacDwdrsww

eas at

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2 Yazdani et al.

J U N E 2 0 1 2 : 6 6 6 – 7 4 Stenting of Saphenous Vein Grafts

669

fractures in the Cypher stents and in the Bx Velocity BMS(which has the identical platform as the Cypher) werelocated in the flexible N-shaped, undulating longitudinalintersinusoidal-ring linker segments, whereas in Taxusstents, factures were observed in the straight longitudinalintercrown linker.

Figure 1. Progression of Atherosclerotic Disease in SVBGs

(A) Within the first year, arterialization and fibrointimal thickening of the veintima. (C) Between the first and third year, formation of a lipid core is observedlumen narrowing is observed. (E) At 5 to 10 years, plaque rupture of a large nAt the coronary anastomosis, fibrointimal growth is commonly observed, wher

Table 2. Patient and Lesion Characteristics of DES- and BMS-TreatedSVBG (>30 Days)

DES (n � 9) BMS (n � 9) p Value

Age, yrs 71 �10 67 �6 0.34

Male 6 (66) 7 (78) 0.62

SVBG duration, yrs 8.0 (6.5–10.0) 9.0 (7.0–12.0) 0.53

Stent duration, days 360 (110–720) 360 (360–900) 0.39

Underlying SVBG disease

FIT 4 (44) 5 (56) 0.57

FA 1 (11) 0 (0)

FA with HNC 4 (44) 4 (44)

Cypher/Taxus 3/6 — —

Stents 2 (1–2) 1 (1–1) 0.13

Stent length, mm 20.0 (14.0–47.5) 13.0 (11.0–26.5) 0.12

Stent fracture 5 (56) 1 (11) 0.045

Thrombosis 4 (44) 0 (0) 0.023

Restenosis 0 (0) 3 (33) 0.058

Malapposition 2 (22) 0 (0) 0.13

Strut NC penetration 4 (44) 4 (44) 0.49

Cause of death

Stent related 4 (44) 3 (33) 0.38

Non–stent-related cardiac 4 (44) 4 (44)

Noncardiac 1 (11) 2 (22)

Values are mean � SD, n (%), and median (interquartile range).

BMS � bare-metal stent(s); DES � drug-eluting stent(s); NC � necrotic core; other abbrevia-

tions as in Table 1.

Morphometric analysis of stented lesions. Lumen, stent, andplaque area measurements were comparable between theBMS and DES groups. Mean neointimal thickness (0.30mm [IQR: 0.20 to 0.48 mm] versus 0.13 mm [IQR: 0.06 to0.16 mm], p � 0.004), neointimal area (4.90 � 2.20 mm2

vs. 2.00 � 0.87 mm2, p � 0.002], and percent area stenosis(47.9 � 20.5% vs. 30.1 � 12.4%, p � 0.041) wereignificantly greater in BMS than in DES, respectivelyTable 3).Histological observations. A total of 216 sections of stented

VBGs were examined with 12 � 3 stent struts per sectionor a total of 785 struts. The underlying lesion characteristicf the stented vein grafts ranged from FIT, FA with aecrotic core, and FA with HNC, with similar distributionsetween DES and BMS cases (Online Tables 1 and 2). InES, fibrin was observed at later time points and was

bserved more frequently around stent struts as comparedith BMS; fibrin was interspersed with inflammatory cells

nd/or surrounded by smooth muscle cells and proteogly-an/collagen matrix (Table 3). The percentage of strutsssociated with inflammatory cells (eosinophils, lympho-ytes, and foreign-body giant cells) was similar between theES and BMS. Luminal neutrophil and eosinophil fociere numerically greater in DES versus BMS, but theifferences were not statistically significant (Table 3). Dis-uption of lipid core by stent struts was observed in alltented SVBG atherosclerotic lesions (DES and BMS ) inhich an underlying FA was present, except in 1 lesionhere a thick fibrous cap of �700 �m was seen. Figures 2 and 3

illustrate necrotic core strut penetration of DES and BMS,respectively. The frequency of stent strut penetration into

s observed. (B) This is followed by foam cell accumulation within the neoin-After 4 to 5 years, hemorrhage into the lipid core and moderate-to-severec core accompanied with hemorrhage often leads to luminal thrombus. (F)herosclerosis is uncommon. SVBG � saphenous vein bypass graft.

graft i. (D)ecroti

pre-existing lipid core was similar between stent groups.

Smealftbil(mmspdpd

pfdcwTrloi

shbpistpqswc

tt

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2

J U N E 2 0 1 2 : 6 6 6 – 7 4

Yazdani et al.

Stenting of Saphenous Vein Grafts

670

Uncovered struts were more common in DES comparedwith BMS, and endothelialization of DES SVBGs wassignificantly less frequent compared with BMS (Table 3).For stents implanted �360 days, uncovered struts weremore frequently observed in samples with strut necrotic corepenetration, with greater frequency in DES as comparedwith BMS, whereas stents without strut necrotic corepenetration, irrespective of stent type, were more likely to becovered (Online Tables 1 and 2).

Evidence of accelerated in-stent atherosclerotic changewithin the neointima was present with equal frequency inBMS (3 of 9 lesions) and DES (4 out of 9 lesions), and wasobserved at 360 days post-implant (Fig. 4). Malappositionof stent struts was only observed in DES SVBGs (2 of 9lesions, both PES). The percentage of struts associated withcalcification was similar between BMS and DES (Table 3).

Discussion

The present study demonstrates that flow-limiting SVBGplaques are characterized by FA with large necrotic coresand are frequently accompanied by plaque hemorrhage anda disrupted fibrous cap. In the treatment of diseased SVBG,DES reduced neointimal thickness as compared with BMS,but this benefit was associated with delayed healing char-acterized by persistent fibrin deposition, uncovered stentstruts, and incomplete endothelialization. Focal stent strutpenetration into lipid core was common and independent ofBMS or DES stent type, and resulted in long-term (�360days) uncovered struts with greater frequency in DES as

Table 3. Pathological and Morphometric Comparison ofDES- and BMS-Stented SVBG

DES (n � 9) BMS (n � 9) p Value

Lumen area, mm2 5.57 � 3.20 6.33 � 3.85 0.65

Stent area, mm2 6.7 (5.4–9.4) 11.1 (6.0–15.5) 0.15

Outer vein area, mm2 15.23 � 6.63 19.13 � 9.42 0.32

NC area, mm2* 2.7 (1.3–4.0) 2.5 (1.0–4.0) 0.95

Plaque area, mm2 7.67 � 3.29 7.91 � 4.97 0.91

Neointimal area, mm2 2.00 � 0.87 4.90 � 2.20 0.002

Stenosis, % 30.05 � 12.44 47.86 � 20.54 0.041

Struts with inflammation, % 20.0 (11.9–54.0) 16.1 (10.1–27.8) 0.72

Struts with calcification, % 6.7 (0–38.0) 17.6 (0–30.5) 0.96

Struts penetrating NC, %* 9.8 (1.7–19.1) 13.0 (4.1–30.5) 0.25

Mean no. ofneutrophils/eosinophils

21.6 (6.3–43.0) 7.0 (4.0–34.5) 0.34

Neointimal thickness, mm 0.13 (0.06–0.16) 0.30 (0.20–0.48) 0.004

Struts with fibrin, % 81.2 (58.3–92.8) 50.0 (30.5–63.9) 0.019

Uncovered struts, % 19.6 (3.3–55.2) 5.1 (0.0–7.3) 0.043

Endothelialization, % 70.0 (60.6–92.5) 98.3 (89.2–100) 0.028

Values are mean � SD or median interquartile range. p values in bold are statistically significant.

*Only performed in lesions with necrotic core.

Abbreviations as in Tables 1 and 2.

compared with BMS. Late stent thrombosis was seen more c

frequently in SVBG DES (4 of 9 lesions) versus BMS (0 of9 lesions), which was almost exclusively seen in vein graftswith underlying large HNCs. Taken together, these find-ings suggest the long-term increase in delayed healing ofDES stents implanted in SVBG is related to the severity ofthe underlying disease and uncovered struts, which result inlate stent thrombosis.Vein graft disease. Compared with native coronary arteries,

VBGs undergo more rapid atherosclerotic lesion develop-ent. Reported rates of vein bypass graft occlusion is 8%

arly after the operation, 13% at 1 year, 20% at 5 years, 41%t 10 years, and 45% at more than 11.5 years (4,25). Ouraboratory and others have shown that within the first yearollowing SVBG implantation, the vein wall is thickened byhe neointimal growth, which has been postulated to occurecause of the veins being exposed to arterial pressure thats 10-fold higher than venous pressure. This hemodynamicoad leads to increased levels of intracellular adhesion moleculeICAM)-1, vascular adhesion molecule (VCAM)-1, andonocyte chemotactic protein (MCP)-1 (26). In animalodels, the magnitude of shear stress across the endothelial

urface of vein grafts has been shown to be inverselyroportional to neointimal growth, suggesting that theifference in flow rates between arterial and venous bedslays an important role in the progression of vein graftisease (27).Over time, the SVBG neointima is infiltrated by macro-

hages, with varying degrees of apoptosis and necrotic coreormation (28). Compared with native coronary arteryisease, SVBG atherosclerotic lesions are more often con-entric and diffuse, with a less well-defined fibrous cap,hich is more vulnerable to rupture and thrombosis (29).he present study demonstrates a high frequency of plaque

upture (32%, 10 of 31 SVBG) of a thin fibrous cap inesions containing large HNCs. It is likely that implantationf stents in SVBGs with thin fibrous caps and a large HNCs an important factor in the healing of DES and BMS.Delayed vein graft healing after DES. The impact of DES onaphenous vein graft healing is poorly understood. Fromuman native coronary arterial autopsy studies, DES haveeen shown to effectively suppress smooth muscle cellroliferation, followed by prolonged fibrin deposition andncomplete endothelialization in noncomplex lesions ob-erved up to 12 months post-DES implantation. By con-rast, endothelialization is nearly complete by 3 to 4 monthsost-BMS placement (30). In the present study, the fre-uency of uncovered stent struts and fibrin deposition wasignificantly higher is SVBGs stented with DES comparedith BMS, which is similar to observations in native

oronary arteries (17).In comparing the histopathology of DES in SVBG (in

he presence of a necrotic core) to native coronary arteries,here was a similar incidence of strut penetration of necrotic

ore and the number of uncovered struts (22). However, a

f thro

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2 Yazdani et al.

J U N E 2 0 1 2 : 6 6 6 – 7 4 Stenting of Saphenous Vein Grafts

671

greater incidence of fibrin was observed in SVBG stentedwith DES as compared with native arteries in patientspresenting with acute coronary syndrome. There was alsogreater delayed healing in SVBG than in native coronaryarteries, likely because of greater delay in endothelializationof vein graft disease. Because the number of cases with DESin SVBG was far too few, it is difficult to compare patientspresenting with acute myocardial infarction in SVBG versusnative coronaries. However, we suspect that there would begreater delayed healing and early neoatherosclerotic changein vein grafts than in native coronary disease.

The data also demonstrate the adverse effects of stentstruts within a necrotic core on stent healing. In DESsamples implanted for �360 days with necrotic core pene-tration, greater incidence of uncovered struts was observedin comparison with DES without strut necrotic core pene-tration, where nearly all struts were completely covered. Asthe presence of uncovered struts highly correlates withendothelialization, neointimal coverage of stent struts couldbe used as a surrogate marker for endothelialization. Un-doubtedly these results correlate directly to the underlyingdisease in SVBG, as strut necrotic core penetration was onlyobserved in grafts with large HNCs. This pattern was

Figure 2. Histological Evidence of Delayed Healing and NC Strut Penetratio

An 84-year-old women received 2 overlapping paclitaxel-eluting Taxus stents ilipid-rich necrotic core (NC) region. The image of the stents at the top showsbox in each panel indicates the location of the enlarged images a, b, and c (rupted by the stent struts (asterisks). (B) In the overlapping region, struts areposed of fibrin. (C) In the distal region, struts are surrounded by a thin layer o

observed both for BMS and DES with incidence of necrotic

core penetration being similar; nevertheless, the number ofuncovered struts was much greater in DES as comparedwith BMS (Online Tables 1 and 2).

Atherosclerotic change was observed following stentingof SVBG in DES as well as BMS, and this change occurredas early as 360 days. These findings are similar to thosepreviously reported by Depre et al. (31) and dissimilar tothose of van Beusekom et al. (32). These differences aremost likely related to overinterpretation of atheroscleroticchange by van Beusekom et al. at early time points (3 to 7months) and likely represented native vein graft atheroscle-rotic disease, as suggested by the authors in their discussion.By contrast, in the study by Depre et al., atheroscleroticchange was observed in 3 of the 12 stent grafts (25%), andthe duration of stent implant was 15, 24, and 24 months,which is similar to our data showing that the earliestatherosclerotic change occurred beyond 12 months.Clinical implications. The present study demonstrates thatDES use for patients with SVBG disease is associated witha greater incidence of uncovered struts, persistence of fibrin,and less endothelialization. Because healing is delayed forlong periods (�1 year), the safety and efficacy of using DEScannot be determined by relatively short-term studies.

ES-Treated SVBGs

ted 2 years before death, that demonstrate stent strut penetration into thecation of the cross sections described in A, B, and C (middle). The yellowm). (A) The fibrous cap (green arrow, bottom) is shown to have been dis-ered but are surrounded by a thin layer of thrombus (Thr) primarily com-mbus. DES � drug-eluting stent(s); SVBG � saphenous vein bypass graft.

n in D

mplanthe lobottouncov

Recently Lee et al. (33) published a meta-analysis of 19

bpaDab

rlSaam

C

focal a

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2

J U N E 2 0 1 2 : 6 6 6 – 7 4

Yazdani et al.

Stenting of Saphenous Vein Grafts

672

studies (2 randomized trials and 17 registries) withfollow-up data ranging from 6 to 48 months. Althoughmost studies with relatively short-term follow up (�2 years)demonstrated a benefit in DES versus BMS (9 of 13studies), longer-term (�2 years) studies were less consistent,with only 3 of 6 studies showing advantages for DESimplantation. In the only randomized study (RRISC [Re-duction of Restenosis In Saphenous vein grafts with Cyphersirolimus-eluting stent] trial) to date, there was a significantreduction of target lesion revascularization in the DESgroup at 6 months versus BMS (5% vs. 22%, p � 0.047),ut this benefit was lost at 3-year follow-up (24% vs. 30%,� 0.55) (14,15). Similarly, Gioia et al. (34) demonstratedmarginal benefit in major adverse cardiac event rates forES at 1 year (10% vs. 16%, p � 0.15); however, major

dverse cardiac event–free survival rates were comparableetween DES and BMS at 2 years (19% vs. 18%, p � 0.9).

These data suggest that late in-stent neointimal growth“catch-up” occurs beyond 2 years in SVBGs treated withDES. Catch-up has been shown in native coronary arteries;

Figure 3. Stent Strut Penetration of Lipid-Rich NC Lesions

A 79-year-old woman received a bare-metal Multi-Link Vision stent 2 years beflipid-rich necrotic core (NC). The image of the stent at the top shows the locapanel indicates the location of the enlarged images a and b (bottom). (A) A t(B) In the distal region, all struts are shown to be covered by neointima with

rates of target vessel revascularization increase for both PES

(Taxus) and SES (Cypher) from 1 to 5 years post-stentimplantation (35). In pathological studies of human post-mortem stented coronary arteries, our laboratory demon-strated a significant increase in neointimal thickness beyond 18months in SES and PES (36). In the current study, because ofthe small number of specimens examined, it is not possible toshow late catch-up in SVBG. Prospective, randomized long-term studies will be needed to further define the risk-benefitprofile of DES implantation in SVBGs.Study limitations. Because this is an autopsy study, theesults may not adequately reflect the vascular responses ofiving patients who receive DES and BMS for treatment ofVBGs. Moreover, the small number of patients in this studynd incomplete clinical data in some subjects did not permit anssessment of the contribution of deployment strategies thatay have contributed to the pathological findings.

onclusions

The pathological characteristics of flow-limiting SVBG

ath, that demonstrates multiple stent struts (asterisk) penetrating into thef the cross sections described in A and B (middle). The black box in eachyer of neointima (green arrow, bottom) is observed over the necrotic core.reas of calcification (black arrow, bottom).

ore detion ohin la

plaques are characterized by FA with large necrotic cores

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2 Yazdani et al.

J U N E 2 0 1 2 : 6 6 6 – 7 4 Stenting of Saphenous Vein Grafts

673

and are frequently accompanied by plaque hemorrhage andor a disrupted fibrous rupture. Stenting of these lesions,regardless of stent type, often results in strut penetration ofthe necrotic core. For DES, penetration of struts intonecrotic core contributes to increased delayed healing, likelybecause of longer duration of drug retention resulting ingreater number of uncovered struts, which may contributeto late thrombotic events. Treatment strategies that reducethe lipid core burden within SVBGs warrant investigationto improve outcomes post-DES implantation.

Reprint requests and correspondence: Dr. Renu Virmani, CV-Path Institute Inc., 19 Firstfield Road, Gaithersburg, Maryland20878. E-mail: rvirmani@cvpath.org.

REFERENCES

1. Hamby RI, Aintablian A, Handler M, et al. Aortocoronary saphenousvein bypass grafts. Long-term patency, morphology and blood flow inpatients with patent grafts early after surgery. Circulation 1979;60:901–9.

2. Virmani R, Atkinson JB, Forman MB. Aortocoronary saphenous veinbypass grafts. Cardiovasc Clin 1988;18:41–62.

3. Laird-Meeter K, van Domburg R, van den Brand MJ, Lubsen J, Bos E,Hugenholtz PG. Incidence, risk, and outcome of reintervention afteraortocoronary bypass surgery. Br Heart J 1987;57:427–35.

4. Fitzgibbon GM, Kafka HP, Leach AJ, Keon WJ, Hooper GD, BurtonJR. Coronary bypass graft fate and patient outcome: angiographicfollow-up of 5,065 grafts related to survival and reoperation in 1,388

Figure 4. Atherosclerotic Change in SVBG Following Stenting

(A and B) A paclitaxel-eluting Taxus stent implanted 5 months before death inlayer (asterisks indicate the stent struts). (C and D) A bare-metal Multi-Link Vrophages within the neointimal layer. SVBG � saphenous vein bypass graft. ThD, respectively.

patients during 25 years. J Am Coll Cardiol 1996;28:616–26.

5. Hong MK, Mehran R, Dangas G, et al. Are we making progress withpercutaneous saphenous vein graft treatment? A comparison of 1990 to1994 and 1995 to 1998 results. J Am Coll Cardiol 2001;38:150–4.

6. de Feyter PJ, van Suylen RJ, de Jaegere PP, Topol EJ, Serruys PW.Balloon angioplasty for the treatment of lesions in saphenous veinbypass grafts. J Am Coll Cardiol 1993;21:1539–49.

7. Serruys PW, de Jaegere P, Kiemeneij F, et al. A comparison ofballoon-expandable-stent implantation with balloon angioplasty inpatients with coronary artery disease. Benestent Study Group. N EnglJ Med 1994;331:489–95.

8. Brener SJ, Ellis SG, Apperson-Hansen C, Leon MB, Topol EJ.Comparison of stenting and balloon angioplasty for narrowings inaortocoronary saphenous vein conduits in place for more than fiveyears. Am J Cardiol 1997;79:13–8.

9. Savage MP, Douglas JS Jr., Fischman DL, et al., for the SaphenousVein De Novo Trial Investigators. Stent placement compared withballoon angioplasty for obstructed coronary bypass grafts. N EnglJ Med 1997;337:740–7.

10. Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versusstandard stents in patients with stenosis in a native coronary artery.N Engl J Med 2003;349:1315–23.

11. Stone GW, Ellis SG, Cox DA, et al. A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med2004;350:221–31.

12. Chu WW, Rha SW, Kuchulakanti PK, et al. Efficacy of sirolimus-eluting stents compared with bare metal stents for saphenous vein graftintervention. Am J Cardiol 2006;97:34–7.

13. Brilakis ES, Saeed B, Banerjee S. Use of drug-eluting stents insaphenous vein aortocoronary bypass graft lesions: a critical appraisal.J Interv Cardiol 2008;21:151–7.

14. Vermeersch P, Agostoni P, Verheye S, et al. Randomized double-blindcomparison of sirolimus-eluting stent versus bare-metal stent implan-tation in diseased saphenous vein grafts: six-month angiographic,intravascular ultrasound, and clinical follow-up of the RRISC trial.

4-year-old women demonstrates foamy macrophages within the neointimaltent implanted 3 years before death in a 66-year-old man shows foamy mac-ted boxes in A and C indicate the location of the enlarged images in B and

an 8ision se dot

J Am Coll Cardiol 2006;48:2423–31.

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 5 , N O . 6 , 2 0 1 2

J U N E 2 0 1 2 : 6 6 6 – 7 4

Yazdani et al.

Stenting of Saphenous Vein Grafts

674

15. Vermeersch P, Agostoni P, Verheye S, et al. Increased late mortalityafter sirolimus-eluting stents versus bare-metal stents in diseasedsaphenous vein grafts: results from the randomized DELAYEDRRISC trial. J Am Coll Cardiol 2007;50:261–7.

16. Goswami NJ, Gaffigan M, Berrio G, et al. Long-term outcomes ofdrug-eluting stents versus bare-metal stents in saphenous vein graftdisease: results from the Prairie “Real World” Stent registry. CatheterCardiovasc Interv 2010;75:93–100.

17. Joner M, Finn AV, Farb A, et al. Pathology of drug-eluting stents inhumans: delayed healing and late thrombotic risk. J Am Coll Cardiol2006;48:193–202.

18. Farb A, Weber DK, Kolodgie FD, Burke AP, Virmani R. Morpho-logical predictors of restenosis after coronary stenting in humans.Circulation 2002;105:2974–80.

19. Kolodgie FD, Virmani R, Burke AP, et al. Pathologic assessment ofthe vulnerable human coronary plaque. Heart 2004;90:1385–91.

20. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points incoronary stent trials: a case for standardized definitions. Circulation2007;115:2344–51.

21. Finn AV, Joner M, Nakazawa G, et al. Pathological correlates of latedrug-eluting stent thrombosis: strut coverage as a marker of endothe-lialization. Circulation 2007;115:2435–41.

22. Nakazawa G, Finn AV, Joner M, et al. Delayed arterial healing andincreased late stent thrombosis at culprit sites after drug-eluting stentplacement for acute myocardial infarction patients: an autopsy study.Circulation 2008;118:1138–45.

23. Nakazawa G, Otsuka F, Nakano M, et al. The pathology of neoath-erosclerosis in human coronary implants bare-metal and drug-elutingstents. J Am Coll Cardiol 2011;57:1314–22.

24. Nakazawa G, Finn AV, Vorpahl M, et al. Incidence and predictors ofdrug-eluting stent fracture in human coronary artery a pathologicanalysis. J Am Coll Cardiol 2009;54:1924–31.

25. Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease:pathogenesis, predisposition, and prevention. Circulation 1998;97:916–31.

26. Zou Y, Dietrich H, Hu Y, Metzler B, Wick G, Xu Q. Mouse modelof venous bypass graft arteriosclerosis. Am J Pathol 1998;153:1301–10.

27. Meyerson SL, Skelly CL, Curi MA, et al. The effects of extremely low

shear stress on cellular proliferation and neointimal thickening in thefailing bypass graft. J Vasc Surg 2001;34:90–7.

28. Butany JW, David TE, Ojha M. Histological and morphometricanalyses of early and late aortocoronary vein grafts and distal anasto-moses. Can J Cardiol 1998;14:671–7.

29. Peykar S, Angiolillo DJ, Bass TA, Costa MA. Saphenous vein graftdisease. Minerva Cardioangiol 2004;52:379–90.

30. Finn AV, Nakazawa G, Joner M, et al. Vascular responses to drugeluting stents: importance of delayed healing. Arterioscler ThrombVasc Biol 2007;27:1500–10.

31. Depre C, Havaux X, Wijns W. Pathology of restenosis in saphenousbypass grafts after long-term stent implantation. Am J Clin Pathol1998;110:378–84.

32. van Beusekom HM, van der Giessen WJ, van Suylen R, Bos E,Bosman FT, Serruys PW. Histology after stenting of human saphenousvein bypass grafts: observations from surgically excised grafts 3 to 320days after stent implantation. J Am Coll Cardiol 1993;21:45–54.

33. Lee MS, Yang T, Kandzari DE, Tobis JM, Liao H, Mahmud E.Comparison by meta-analysis of drug-eluting stents and bare metalstents for saphenous vein graft intervention. Am J Cardiol 2010;105:1076–82.

34. Gioia G, Benassi A, Mohendra R, Chowdhury K, Masood I, MatthaiW. Lack of clinical long-term benefit with the use of a drug elutingstent compared to use of a bare metal stent in saphenous vein grafts.Catheter Cardiovasc Interv 2008;72:13–20.

35. Park KW, Kim CH, Lee HY, et al. Does “late catch-up” exist indrug-eluting stents: insights from a serial quantitative coronary angiog-raphy analysis of sirolimus versus paclitaxel-eluting stents. Am Heart J2010;159:446-53.e3.

36. Nakazawa G, Finn AV, Vorpahl M, Ladich E, Kolodgie FD,Virmani R. Coronary responses and differential mechanisms of latestent thrombosis attributed to first-generation sirolimus- andpaclitaxel-eluting stents. J Am Coll Cardiol 2011;57:390–8.

Key Words: bare-metal stent(s) � drug-eluting stent(s) �pathology � saphenous vein bypass graft.

APPENDIX

For supplementary figures and tables, please see the online version of this

article.