From the Department of Cardiology, Cleveland Clinic Foundation,
Cleveland, Ohio, and the Department of Interventional Cardiology, Thorax
Center, Erasmus University, Rotterdam, Netherlands.
Correspondence to Eric J. Topol, MD, Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195.
In more than 20 years since the first
percutaneous coronary
revascularization procedures, the field of
interventional cardiology has proliferated beyond all
expectations. Now more than 1 million procedures are performed
worldwide each year. Stenting has revolutionized the field, which
previously relied on balloon dilatation in the majority of patients.
With
Platelets
Proximate Cause of Stent Thrombosis
Evidence for causal white thrombus came from platelet flow
cytometry data, indicating the increased expression of the platelet
glycoprotein IIb/IIIa receptor in patients who developed
this complication of stenting.13 Rather than
abnormalities in markers of coagulation activation, such as the cleaved
fragment of prothrombin (F1.2) or thrombin-antithrombin complexes, a
series of consecutive patients demonstrated that the dominant
biochemical abnormality was related to platelet activation.
Angioscopic data from patients with subacute stent thrombosis
showed white (platelet-rich) thrombus as the
culprit,14 and reports of failure of
plasminogen activators to achieve lysis of
stent thrombosis15 corroborated the sense that
the pathogenesis of this entity was largely attributable to
platelet aggregation. Clinical trials were mounted to test the
value of a platelet-directed strategy of aspirin and ticlopidine
compared with the full anticoagulation approach of prolonged heparin
and coumarin.16 17 Furthermore, trials compared
aspirin as a sole antiplatelet agent with the combination of
aspirin and the ADP receptor antagonist
ticlopidine.17 18 19 20 These trials are summarized in
Table 1
Recently, ticlopidine has been associated with thrombocytopenic
thrombotic purpura.22 23 A new ADP receptor
antagonist, clopidogrel, has thus far not been shown to
induce this serious microangiopathic disorder or been implicated in
bone marrow toxicity. Clopidogrel is currently being assessed in
coronary stent trials for comparative efficacy and safety
compared with ticlopidine.
Development of IIb/IIIa Inhibitors
Six large placebo-controlled, double-blind trials were undertaken to
test IIb/IIIa antagonists in patients undergoing PTCA, as
summarized in Table 2
The subsequent 5 trials confirmed and extended the EPIC findings in
many ways.27 28 29 30 31 First, each trial showed benefit
of IIb/IIIa inhibition (Figure 1
Stents and IIb/IIIa Inhibitors
The IIb/IIIa trials in interventional cardiology,
interestingly, have consistently reduced the need for emergency
or "bailout" stenting33 34 (Figure 3
Restenosis
As shown in Figure 4
Periprocedural MI
Use in Acute MI
In the recently completed RAPPORT trial,41 of 483
patients undergoing PTCA for acute MI, the IIb/IIIa
inhibitor group had a >40% reduction in death, MI, or
urgent revascularization at 30 days (Figure 5
It is thus likely that a pivotal component of future management of MI
patients will involve more potent antiplatelet therapy. The
long-standing debate as to whether to perform catheter-based
reperfusion or use pharmacological therapy could be preempted by a
pharmacological strategy that markedly facilitates (rather than
compromises) acute-phase intervention. Because one of the most
important features of acute intervention is the "door-to-balloon"
time,57 58 the use of a platelet-lysis
approach to restore coronary blood flow while the patient is
being transported to cardiac catheterization has
important infarct-size limitation and practical appeal. Furthermore,
the enhanced stability of the infarct vessel that has been passivated
by the use IIb/IIIa blockade opens the potential for consideration of
very early hospital discharge. Thus, IIb/IIIa blockade has set the
stage for a veritable revolution in our care of the patient with acute
MI eligible for reperfusion therapy.
Antirestenosis Strategies
The focus of the treatment of restenosis over the last 2
decades has been through the application of pharmacologically active
agents and mechanical approaches using a host of different devices.
Unfortunately, this frequent and costly complication of
percutaneous revascularization
techniques has proved refractory to all such therapies. The
restenosis process is characterized by neointimal
proliferation, which involves the migration of vascular smooth muscle
cells from the media to the intima and their subsequent proliferation.
The inciting stimuli involved in restenosis include disruption
of the endothelial barrier layer, mechanical factors
that disrupt the medial smooth muscle layer and serve as stimuli for
smooth muscle cell proliferation and migration, and the contact of this
disrupted layer with circulating blood factors and mitogens that serve
as further stimuli to neointimal formation. Vascular injury
sets into motion a cascade of events that results in the final
hyperplastic response shown by the neointima. Early in our
understanding of the pathophysiological processes
involved in restenosis, attention was concentrated on factors
that interact with smooth muscle cells through cell surface receptors.
These include such compounds as thrombin, platelet-derived growth
factor, angiotensin II, interleukin 1, insulin growth
factor-l, basic fibroblast growth factor, and a whole host of other
mitogenic factors. Initial treatments for
restenosis targeted these receptors with pharmacological agents
in an attempt to inhibit their effects. It soon became apparent,
however, that none of these stimuli work through a unique or final
common pathway. Instead, these signals interact at an intracellular
level through redundant second-messenger systems, which confers
redundancy to the system. Ultimately, these second-messenger systems
converge on a final common pathway at the cellular DNA level known as
the cell cycle.
The life cycle of a normal cell can be considered in 5 different cell
phases. G0 (G=gap) is the quiescent state in
which the cell is biologically active but is neither actively dividing
or replicating. Under appropriate stimuli, the cell can then enter the
G1, or interphase, in which biosynthetic
activities of the cell prepare it to enter into the next phase of the
cell cycle, or S phase. The S phase begins when DNA synthesis starts
and ends when the DNA content of the nucleus has doubled and the
chromosomes have replicated. The S phase is followed by
G2, which ends when mitosis starts, signaling the
start of the M phase. Activation of the cell cycle is responsible for
both normal physiological growth and cell division,
as well as for pathophysiological processes such as
restenosis.
A large number of genes are involved in the control of cell cycle
progression in eukaryotic cells. They can be divided into
early G1 genes (such as c-fos
c-myc, etc), late G1 genes
(c-myb, Rb, etc), genes involved in
G1/S transition (cdc/ckd
kinases, etc), S-phase genes involved in DNA synthesis (DNA
polymerases, proliferating cell nuclear antigen, etc), genes involved
in G2/M transition (cdc/ckd
kinases, etc), and genes involved in mitosis (cytoskeletal proteins,
mitosis-specific kinases, etc). In theory, suppression of any 1 of
these genes will lead to interruption of cell cycle progression, a
strategy that is being explored for the prevention of
restenosis as we approach the new millennium. Two different
tactics are currently being tested. Brachytherapy bombards the nuclear
material with ionizing radiation, disrupting the templates for the
production of cell cycle regulatory proteins. With the correct
radiation dose, the target smooth muscle cell remains viable but unable
to replicate. The other approach depends on a biological attack on the
cell regulatory machinery using gene therapy technology for the
introduction of foreign-fragment nuclear material to halt the
pathophysiological processes typical of
restenosis.
A number of different approaches using the basic techniques of gene
therapy are currently being tested for the prevention of
neointimal hyperplasia. The most straightforward involve
the transfer of a gene directly into the proliferating smooth muscle
cells. This gene can encode for a cytotoxic protein (eg, herpes virus
thymidine kinase), a cell cycle–inhibitory protein (eg,
p53, p21, Rb, cdk), angiogenic proteins (eg, vascular
endothelial growth factor [VEGF], angiogenin, basic
fibroblast growth factor), or proteins with vasodilatory,
antithrombotic, or antiproliferative properties (eg, nitric oxide
synthase, cox-l, plasminogen activator
inhibitor type 1). This kind of approach has been shown to
be effective in several animal models of restenosis. With
adenovirus used as a vector, herpes virus thymidine kinase has been
successfully transfected into vascular smooth muscle cells at the site
of balloon injury in the femoral arteries of
swine.59 This was followed by systemic
administration of the nucleoside analogue ganciclovir, which in the
presence of the foreign thymidine kinase can be incorporated into the
cellular DNA. With this technique, both smooth muscle cell
proliferation and luminal narrowing were shown to be
inhibited.59 Adenovirus-mediated transfer of the
retinoblastoma (Rb) gene, whose protein product inhibits
cell cycle progression, has also been shown to be successful in a
similar animal model.60 Transfer of the gene
coding for nitric oxide synthase using both a protein/liposome hybrid
vector61 and an adenovirus
vector,62 as well as of angiotensin
II type 2 receptor63 and the thrombin
inhibitor hirudin64 using adenovirus
vectors, has also been shown to be effective for the prevention of
restenosis in the rat carotid model of vessel injury.
Other targets being investigated for the prevention of
restenosis are the matrix metalloproteinases (MMPs). Smooth
muscle cell migration and proliferation are normally hindered by the
inhibitory constraints of the extracellular
matrix.65 66 Matrix remodeling by
matrix-degrading metalloproteinases is therefore essential for smooth
muscle proliferation and migration into the intima. Synthetic
inhibitors of MMPs have been shown to inhibit smooth muscle
migration and proliferation in several in vitro
systems.66 67 68 69 70 There are several naturally
occurring tissue inhibitors of metalloproteinases (TIMPs)
that can inhibit their activity. To date, 3 TIMPs have been well
characterized (TIMP-1, -2, and -3),71 and a
fourth has been cloned.72 The identification of
these inhibitors has led to the idea that overexpression of
TIMPs may have potential as a gene therapy approach for the prevention
of neointimal formation. Support for TIMP gene therapy
comes from recent work by George and colleagues73
at the Bristol Heart Institute. In human saphenous vein in tissue
culture, these investigators were able to demonstrate that highly
localized overexpression of TIMP-1 after recombinant
adenovirus-mediated transfection markedly inhibited
neointimal formation. These results not only provide proof
of principle for ex vivo gene therapy with TIMP-1 for the reduction of
neointimal formation in vein grafts but also suggest an in
vivo application at the time of intervention for the prevention of
coronary artery restenosis.
Another "nuclear weapon" being developed for the prevention of
restenosis involves antisense oligonucleotide
technology. Antisense oligodeoxynucleotides are short
pieces of DNA with sequences that are complementary to specific regions
of messenger RNA. The major mechanisms of action of these compounds are
through the sequence-specific interaction with messenger RNA, although
sequence-specific and sequence-nonspecific effects have also been
demonstrated. On binding to the target, the antisense compound
sterically inhibits the interaction of ribosomes with the messenger
RNA. Another mechanism of action, which may be as important as the
steric inhibition of ribosome binding, is a consequence of the DNA/RNA
hybrid being more susceptible to degradation by intracellular RNAse
than single-stranded messenger RNA. This results in an increased
clearance of target mRNA from the cell. In principle, any gene may be
selected for antisense suppression; inhibition of certain genes will
certainly be more biologically effective. Important in this regard are
the abundance of the messenger RNA, the half-life of the protein
product, and the existence of redundancy within the cell such that
other proteins are capable of performing functions similar to that
which is targeted. Given these considerations, it is not surprising
that most of the attention of antisense technology has been focused on
the short-lived regulators of the final common pathway of
mitogenic stimuli, the cell cycle. Inhibition of the
production of several of the mediators of the cell cycle with
antisense oligonucleotides has been shown to be
effective for the prevention of restenosis in several different
animal models of vascular injury (Table 5
With our increased understanding of vascular molecular biology, more
sophisticated approaches that use the basics of gene therapy are being
investigated. For instance, cell-based vascular gene-delivery
techniques are now being explored as means to provide biologically
relevant amounts of therapeutic agents to the site of vessel damage.
This strategy involves the isolation of autologous
endothelial or smooth muscle cells, ex vivo gene
transfer, followed by the reintroduction of the genetically modified
cells back to the region selected for therapy. The disadvantages of
this approach include the requirement to isolate and modify cells from
each patient, with the consequent delay in therapy, and the failure to
date, with few exceptions, to show a relevant expression of recombinant
protein in vivo. To circumvent these problems, the laboratory of one of
the authors (P.W.S.) is focusing on the xenotransplantation of
genetically modified endothelial cells for the
prevention of restenosis. For this purpose, animals have been
engineered that are doubly transgenic in that they carry not only a
human gene that can produce an agent for the prevention of
restenosis but also gene coding for human cell surface markers
so that they may be effectively xenotransplanted into humans and
therefore provide a local source of bioactive compounds. These animals
provide a limitless supply of endothelial cells
producing controllable levels of active compound, which can be used for
xenotransplantation at the site of vessel injury in humans. These
foreign cells act as a kind of Trojan horse, graciously accepted as
self by the host organism but capable of modifying the
pathophysiological response to vessel damage
typified by the process of restenosis. Once implanted, the
production of the bioactive compound is under exogenous control
by means of "designer" genes coding for modified cell-surface
receptors that are introduced. In this system, interaction of an
orally administered compound with the modified cell receptor can switch
on the transgene, whereas in its absence the transgene remains dormant.
The feasibility of this type of approach has been demonstrated in other
animal species, and it shows great potential for application to humans.
The applicability of this type of therapeutic delivery system to other
pathophysiological conditions and the shortage of
tissue for organ transplantation are the fuel for continued high
research activity in this direction.
As our understanding of the molecular control of cellular function
increases and our knowledge of the pathophysiology of
restenosis expands, new types of gene therapy approaches will
be developed for the treatment of this iatrogenic complication. An area
of active investigation is the development of new techniques for the
introduction of foreign DNA into whole cells. All of the currently
available viral and nonviral vectors have significant limitations to
their use. The introduction of an effective, nonimmunogenic delivery
vehicle is on the horizon. As the last decade was the era of mechanical
and pharmacological approaches to the prevention of restenosis,
developments in the coming decade will be aimed at manipulating the
genetic material of the proliferating cell.
Stents
The use of coronary stent implantation as a primary
treatment modality in interventional cardiology has
increased at a staggering rate. In most interventional centers, stents
are used in
Are They All Equivalent?
Another factor that must be considered when assessing stent design is
the mode of delivery: self-expanding versus balloon-expandable. The
currently available self-expanding stents are configured as a slotted
tube (Radius) or as a wire mesh (Wallstent) and are composed of either
nitinol (Radius) or stainless steel (Wallstent). Although different in
configuration and metal composition, they have in common a continued
expansion after deployment. The nitinol device will continue to expand
to its nominal programmed diameter, whereas the Wallstent expands to
the point at which tissue forces overcome the radial forces of the
expanding stent.103 104 It must be kept in mind,
however, that excessive oversizing of the Wallstent (>0.7 mm
larger than the reference diameter) has been shown to be a powerful
predictor of subsequent restenosis.105
Further studies are necessary to better define the particular
advantages and disadvantages of the 2 different modes of delivery.
To resolve the issue of whether stent configuration plays a major
role in determining long-term outcome, large randomized trials are
currently under way comparing various stent designs "head to head."
The equivalency design of these trials in simple "BElgian NEtherlands
STENT study [BENESTENT]-like" lesions is predicated on showing
similarity in safety and efficacy between the test stent and the
Palmaz-Schatz stent, which serves as the standard with which all others
are compared. These trials were not designed to test for subtle and
perhaps clinically unimportant differences between stent designs,
because an unreasonably large sample size would be required. Three
trials have been completed. In all of these, stents were implanted in
noncomplex lesions with a length of <25 mm in native
coronary vessels. In addition to clinical end points,
angiography was obtained in a subset of the patients, and all trials
were powered to detect a 0.2-mm difference in minimal luminal diameter
at 6 to 9 months after stent implantation by quantitative
coronary arteriography. The ACS stENT Clinical Equivalence in
de Novo Lesions (ACSENT) trial included 1040 patients randomized to
either Palmaz-Schatz or Multi-Link stent implantation. Statistical
equivalence in both the late clinical and angiographic (Table 6
Custom-Designed Stents
Bifurcation Lesions
CR Bard Inc is testing a true bifurcation stent. The Bard Bifurcation
stent is shaped like a Y and is mounted on 2 balloons (Figure 7B
Side Branch
Ostial Lesions
Aneurysms or Perforations
A novel approach to stent customization for the treatment of vessel
rupture and aneurysms has been developed by Drs Stefanadis and
Toutouzas.106 Their approach involves passivation
of the stent surface through the application of a segment of autologous
vascular tissue. The technique uses a segment of cephalic vein or ulnar
artery to cover the stent. The results of implantation of a device with
a vein graft covering only the external surface of the stent have been
reported both for elective indications106 107 108 109
and in the setting of an acute MI.108 109 110 Radial
artery–covered stents have also been successfully implanted in
saphenous vein bypass grafts,111 112 as well as
in both the body and ostia of native coronary
vessels.112 Further studies are necessary to
clarify the potential of this technique.
Radioactive Stents
Hehrlein and colleagues116 117 were the first to
describe the use of radioactive stents, which they implanted in
nondiseased rabbit iliac arteries. The stainless steel stents were made
radioactive by ion bombardment in a cyclotron and emitted both
Laird et al118 also examined the effects of a
radioactive coil stent. They first ion-implanted the nonradioactive
element 31P beneath the surface of the stent. The
stents were then made radioactive by exposure to neutron irradiation,
which converts a fraction of the 31P atoms to
32P, a pure ß-particle emitter. This technique
results in an even distribution of 32P within the
stent, which ensures homogeneous distribution of
ß-particle irradiation from the stent. This technique, however,
generates other short-lived radioisotopes. Intraluminal exposure for 28
days to these radioactive stents, with an initial activity of 0.014
mCi, caused a significant reduction in neointimal area and
percent area stenosis compared with the effects of
nonradioactive stents.
The efficacy of a relatively low-dose, pure ß-emitting stent for the
inhibition of intimal hyperplasia was first demonstrated by Hehrlein
and colleagues.119 32P,
produced by neutron bombardment, was ionized and ion-implanted beneath
the outer surface of titanium-nickel stents. 32P
has several characteristics. 32P-emitting stents
with activities of 4 and 13 mCi were implanted in rabbit iliac
arteries, and histomorphometry was performed at 4 and 12
weeks.119 At 4 weeks, both groups showed
significant reductions in neointimal formation, whereas at
12 weeks, only the group receiving the highest radiation dose showed a
significant reduction in neointima compared with
nonradioactive stents.
With a similar radioactive stent, the neointimal responses
to implantation in a porcine coronary restenosis model
were examined in stents with activities from 0.15 to 23 mCi.
Neointimal formation was seen to be reduced 28 days after
the implantation of low-activity (0.15 to 0.5 mCi) and high-activity (3
to 23 mCi) stents, but increased neointimal formation was
observed with stents of 1-mCi initial activity. These results highlight
the complexity of the response of the vascular wall to ionizing
radiations. The stent used in this trial has subsequently become known
as the Fischell IsoStent (IsoStent/Cordis Corp, a Johnson & Johnson
Interventional Systems Co.120 It is a stainless
steel Palmaz-Schatz stent that has been modified to be
ß-particle–emitting by ion implantation as described above. Prompted
by the encouraging results of ß-particle–emitting stents on
neointimal hyperplasia in animal
models,117 118 119 121 122 a multicenter pilot study
examining the feasibility and safety of the implantation of 1-mCi
Palmaz-Schatz stents has been completed, and the larger randomized
IsoStent for Restenosis Intervention Study (IRIS) trial is
under way.
Proton activation of nitinol produces the predominantly ß-emitting
isotope 48V. The Act-One stent (Progressive
Angioplasty Systems Inc), the predecessor of the Paragon stent, has
been made radioactive through proton activation and tested in pig
coronary arteries.123 Radioactive Act-One
stents with 1.5 mCi of 48V activity had no effect
on lumen narrowing or vessel histology, whereas 10-mCi
48V stents inhibited neointimal
thickening compared with nonradioactive stented control segments.
Further studies are necessary to assess the effectiveness of
radioactive nitinol stents for the prevention of
restenosis.
Stent Coatings
Another commercially available coated device is the divYsio stent
(Biocompatibles Ltd), which is phosphorylcholine-coated.
Phosphorylcholine is the major phospholipid component of biological
membranes. On the basis of promising results in
vitro148 and in vivo in animal
models,149 150 151 it is anticipated that these
coated devices will behave as intact tissue elements, a form of
biomimicry, and result in a reduction in the incidence of subacute
occlusion and an improvement in the long-term patency rates of treated
segments. These stents are now being evaluated in clinical trials in
Europe for their ability to reduce the incidence of subacute
occlusion and improve the long-term outcome in stented coronary
segments.
Fibrin coating of intravascular stents has been proposed as a means of
passivating the stent surface and providing a platform for the
recolonization of endothelial
cells.138 Fibrin-coated Palmaz-Schatz stents were
shown to be free of thrombus and foreign body reaction when examined 8
weeks after implantation in dog peripheral
arteries.139 This was compared with the 45%
incidence of thrombosis seen with the implantation of naked stents.
More notable was the finding of endothelialization of
96% of the surface of the fibrin-coated stents, whereas the uncoated
controls were covered with endothelial cells over only
18% of their surface. Similar results were seen with implantation in
pig coronary arteries. In this model, no significant
foreign-body, giant-cell, or inflammatory reaction was seen up to 1
year after stent implantation.152
Polymeric coating of the stent in situ has also been shown to be
feasible, a technique referred to as "gel
paving."142 In a recently reported study, the
application of polyethylene-glycol-lactide hydrogel polymers to the
surface of Palmaz stents implanted in the porcine femoral artery
model has been described.153 The applied polymer
is then photopolymerized in situ to form a short-term, semipermeable
barrier. Stented segments treated in such a manner showed less gross
thrombosis, less microscopic platelet adherence, and enhanced
vessel patency compared with control stented segments. Many more animal
data must be collected before this type of technology can be considered
for clinical application.
As a result of their long residence times, attention has become focused
on endovascular stents as a reservoir for prolonged local drug
administration for the prevention of restenosis. This can be
done by coating metallic stents with controlled-release matrices or
incorporating a pharmacologically active compound into a polymeric
stent or a polymer-metal composite stent.
Drug-polymer composites are referred to as monolithic matrices. When
nondegradable matrices are used, drugs are delivered through sustained
release by way of particle dissolution and diffusion through the
cavitating network of the matrix. Extended drug release is possible
through this approach, with formulations with release duration from
hours to decades having been reported. Biodegradable polymer systems
have also been used to formulate drug-delivery matrices. Biodegradable
polymer matrices provide sustained delivery of pharmacological agents
both by drug dissolution and by matrix degradation in vivo, leading to
release of entrapped agents. The coating of a pharmaceutical stent with
a biodegradable polymer also offers the attractive possibility that the
drug-polymer system could disappear after a desired period of drug
release.
Several candidate drugs for stent coatings have been considered.
Undergoing clinical assessment is an InFlow stent (InFlow Dynamics AG)
coated with a polylactic acid carrier containing 5%
polyethylene-glycol-hirudin and 1% prostaglandin
I2 analog (Iloprost). In vitro analysis
demonstrated favorable degradation properties of the carrier and
time-release characteristics of the incorporated antithrombotic and
platelet-inhibiting drugs.154
Analysis of the hirudin- and Iloprost-eluting stents tested
during stasis in a human shunt model demonstrated a significant effect
on both platelet activation and blood
coagulation,155 and when implanted in sheep
coronary arteries, they have been shown to exhibit a favorable
effect on neointimal formation.156
Another carrier/active agent system that appears promising is a
cellulose polymer with passively adsorbed glycoprotein
IIb/IIIa receptor antibody.140 141 157 Active
compound elutes from the stents in an exponential manner, with 48% of
the bound agent eluted at 12 days when studied in
vitro.158 When investigated in a rabbit iliac
artery model, antibody to glycoprotein IIb/IIIa eluted from
cellulose-polymer–coated stents significantly reduced platelet
aggregation in the stent microenvironment, reduced thrombus formation,
improved blood flow and arterial patency rates, and
inhibited cyclic blood flow variation.157
The use of gene therapy in conjunction with a pharmaceutical delivery
stent could involve the transfer of a desired gene from the stent
coating to the cells of the arterial wall. This should
result in the expression and synthesis of a desirable product by
the cells of the arterial wall. This approach would involve
the incorporation of DNA or a viral vector into a polymeric matrix
system under conditions that would facilitate cellular uptake and
translation of the DNA. Another important possible strategy for a
pharmaceutical stent approach might involve the incorporation of
antisense oligonucleotides into an appropriate
polymeric matrix.159 160
Interest in the development of a suitable biodegradable stent with
pharmacologically active agents incorporated into the polymeric matrix,
once a very active area of research, has waned considerably. To be
effective, a drug-releasing biodegradable stent must be biocompatible,
it must not cause an inflammatory reaction, and the breakdown
products must be nontoxic. One such biodegradable device, however,
that warrants mention is the Duke Biodegradable Stent, which is made
from a special form of poly-L-lactide capable of
incorporating pharmacologically active agents.161
Both self-expanding and balloon-expandable versions of the Duke stent
have been designed and tested in animals,162 with
promising results.
The seeding of intravascular stents with endothelial
cells to passivate the stent surface has been an area of ongoing
research for >9 years. Both self-expanding163
and balloon-expandable164 stents have been
successfully seeded with endothelial cells and have
been shown to retain a significant number of viable cells after
deployment in vitro.163 164 165 Using autologous
endothelial cells derived from sheep saphenous veins, a
group at the National Heart, Lung, and Blood Institute (NHLBI) has
successfully seeded metallic stents and implanted the stents into the
femoral arteries of the donor animals.166 The
transplanted endothelial cells could be detected in 6
of 9 animals treated in this manner 10 days after stent implantation.
Scott and colleagues167 also were successful in
identifying endothelial cells on seeded stents 3 hours
after intracoronary implantation in pigs. These investigators
used immortalized human microvascular cells that retain the phenotypic
characteristics of endothelial cells after >50
passages.168 The relevance of continued
investment into the development of endothelium-covered
stents may lie in the future possibility of seeding the stents with
genetically modified endothelial cells capable of
producing compounds for the treatment of restenosis.
Future Advances
Angiogenesis and Transmyocardial Revascularization
Perhaps the most intriguing and exciting frontier in
interventional cardiology is the ability to induce the
growth of new blood vessels, a radically different therapeutic
orientation compared with bypassing or deblocking diseased arteries.
Several approaches have undergone pilot studies and have entered
clinical trials involving either local angiogenic polypeptide growth
factor administration, transfection of angiogenic genes, or the
development of small channels in the endocardium to promote the passage
of blood from the left ventricular cavity into the
myocardium. The principal mechanism of improved myocardial
blood flow for the latter approach, known as transmyocardial
revascularization, is most likely related to
stimulating angiogenesis, because the holes created by the laser (or
other means) become fibrotic and occlude within weeks after the
procedure.168 169 170
Therapeutic angiogenesis is undergoing intensive early clinical
investigation. Isner et al171 172 transfected the
peripheral arteries with the cDNA for
VEGF165 via a hydrogel-coated balloon catheter.
This was successful in a few patients for improving
peripheral arterial ischemia and served
as proof of concept; then the same group initiated intramuscular
injection of the VEGF165 naked human plasmid cDNA
in 9 patients (10 ischemic limbs). In this initial series of
intramuscular (rather than intra-arterial) injection of an
angiogenic gene, 8 of 10 limbs had improved blood flow demonstrated by
MRI, and there was an overall 30% increase in the ankle-brachial
index, a key measure of ameliorated limb blood supply. Patients were
excluded for any known malignancy, diabetic
retinopathy, or other conditions that might be
associated with untoward sequelae from "inappropriate"
angiogenesis.
More recently, the initial forays into coronary angiogenesis
have been reported. Schumacher et al,173 from
Freiburg, Germany, published the results of a randomized trial of 40
patients who had intramyocardial injection of acidic fibroblast growth
factor (FGF-1) into the distal anterior wall at the time of internal
mammary artery bypass graft surgery. Within 12 weeks, the 20 patients
randomized to active FGF-1 had digital angiographic evidence of
neovascularization.173 During 3 years of
follow-up, the ejection fraction improved from 50.3% to 63.8% in the
treated group compared with 51.5% to 59.4% in the control
(heat-inactivated FGF-1) group. No accelerated
atherosclerosis or other untoward events were
found.
In a parallel clinical investigation, intramyocardial injection of the
gene for VEGF121 was initiated using an
adenoviral vector at the time of bypass
surgery.174 Like the clinical studies of
Shumacher and colleagues, there was evidence from this preclinical
porcine model of Mack et al175 of extensive
neovascularization at 4 weeks after VEGF transfection.
Henry and colleagues176 presented their
findings on intracoronary VEGF administration in a small
dose-finding pilot trial of 15 patients, which showed marked
improvement in perfusion in 7 patients (47%) by thallium
scintigraphy and improved collateralization in 5 of 7
patients who underwent follow-up angiography. A larger follow-up study
that combines intracoronary and subsequent
intravenous VEGF has just been initiated. Laham and
associates177 used epicardial microcapsules with
basic fibroblast growth factor in 8 patients at the time of bypass
surgery, 4 of whom had enhanced perfusion by nuclear scans and MRI.
It was not fully anticipated that a single injection of either
angiogenic protein or gene would produce prolonged neovascularization.
This raises the question as to whether there is upregulation of
angiogenic factors or their receptors once the stimulus is
initiated.178 The process of new blood vessel
formation is such a complex and well-orchestrated series of events,
requiring a multitude of growth factors and regulatory proteins, that
it is probably naive to think that a 1-time application of a single
growth factor will yield optimal angiogenesis. The gene transfer
approach may be preferred over the protein with regard to potential for
durability, but with the adenoviral vector, issues including gene
regulation over time and the potential for induction of a host
inflammatory response are poorly understood. The naked cDNA plasmid
strategy, applied for the peripheral arterial
insufficiency indication, is fraught with low efficiency of transfer
and probably more transient gene expression.
Although the angiogenesis can be demonstrated by imaging modalities, it
is uncertain whether it will lead to normalization of perfusion or
complete mitigation of provokable ischemia. It remains to be
seen whether the collateral supply will be enhanced for long-term
follow-up, but the functional data provided by Schumacher's group are,
at the very least, encouraging. Still, the unknown effects of
accelerating atherosclerosis or facilitation of a
latent malignancy need to be carefully assessed. Notwithstanding these
concerns, large-scale clinical investigation for coronary
angiogenesis with either the protein or gene for specific growth
factors is warranted.
Substantial clinical pilot studies of transluminal myocardial
revascularization (TMR) have suggested the
potential for relief of angina.179 180 181 182 183 184 185 Three
clinical trials have been completed with TMR performed at the time of
open-heart surgery. With the CO2 laser, 198
patients were randomized to TMR or initial medical
therapy.186 Holmium laser was also the subject of
a randomized trial that enrolled 162 patients.187
The results of these 2 trials are remarkably concordant with reduction
of angina, improvement of functional status, and in select patients
with ancillary nuclear imaging, improved perfusion. As summarized in
Figure 8
The encouraging results in the initial surgical trials have led to
clinical investigation with the use of holmium or excimer laser via the
percutaneous approach.193 194 To
date, no randomized trial data are available, but pilot studies
demonstrate feasibility and apparent safety in the first few hundred
patients treated. Clearly, the ability to direct the laser to the
ischemic myocardial bed is lessened with the
percutaneous approach, but new catheter-based sensor
systems that are nonfluoroscopic and use electroanatomic mapping are
being investigated to improve on the precision of where the holes are
applied.195 196 Because the dominant mechanism
may relate to stimulation of angiogenesis, it is not clear that
precision is vital. The damage to the myocardium that is
anticipated in the form of non–ST-elevation MI and its long-term
effects has not been amply studied. A percutaneous
strategy has obvious advantages, because it will allow more ready
application to patients who are truly inoperable or can be used as an
adjunct in patients with a chronic total occlusion that is not amenable
to a routine epicardial strategy. Nevertheless, TMR will need
considerable study and comparison with pharmacological or genetic forms
of angiogenesis. Even the concept of injected angiogenic factors at the
time the holes are created has been subjected to experimental study
and, if shown clinically to be synergistic, may someday prove to be a
useful combined strategy.197
Another highly provocative strategy is a tissue-engineered
blood vessel. L'Heureux and associates,198 in
Quebec, have demonstrated the feasibility of cloning new arteries via
cultured human smooth muscle cells, endothelial cells,
and fibroblasts. Although this seems futuristic, it is possible that
larger, bioengineered vessels cloned from patients who have significant
ischemia may be a viable alternative in the next millennium to
the conventional epicardial or emerging angiogenesis interventions of
today.
Overall Perspective
References
© 1998 American Heart Association, Inc.
Clinical Cardiology: New Frontiers
Frontiers in Interventional Cardiology
Key Words: revascularization • stents • platelets • restenosis • angiogenesis
50% of patients now undergoing stent implantation, the
groundwork is laid for further important advances. In this article, we
discuss the 4 most important new advances in the field of
interventional cardiology: platelet inhibition,
prevention of restenosis, stent evolution, and
angiogenesis.
Considerable progress in the field of percutaneous
coronary revascularization has been made by
appreciation of the pivotal role of platelets and the development
of new therapies directed against platelet aggregation. The entire
field of stenting was catapulted forward with the realization that
subacute thrombosis was predominantly attributable to a
platelet thrombus. Since the beginning of coronary stenting
in 1986, thromboprophylaxis consisted of full doses of heparin and
coumarin, along with aspirin and
dipyridamole.1 2 This regimen
resulted in the necessity for prolonged hospital stays to achieve
therapeutic anticoagulation and in excessive bleeding complications and
was still ineffective in preventing stent thrombosis in at least 4% to
5% of patients.2 3 4 5 6 The prevailing belief was
that stent thrombosis, which occurred primarily several days after the
index procedure, was due to a fibrin-rich (red)
thrombus.7 8 9 Many of the cases of stent
thrombosis were linked to suboptimal
anticoagulation.8 9 The complication of stent
thrombosis is dreaded, because it may result in myocardial infarction
(MI) or death.10 The possibility that stent
thrombosis was precipitated by inadequate expansion of the stent was
explored11 12 independently of the search for a
better pharmacological protective strategy. 
. The benefit of an
antiplatelet approach over the traditional strategy was striking,
as first shown in the Munich Intracoronary Stenting and
Antithrombotic Regimen (ISAR) trial and later confirmed in the Stent
Anticoagulation Regimen Study (STARS)
trial.16 17 These findings, coupled with the
benefit of an intensified dual antiplatelet inhibitor
approach as opposed to aspirin alone,17 18 19 20
provide a very strong foundation for platelet thrombus to be
regarded as the principal pathogenic factor in stent thrombosis. The
incidence of subacute stent thrombosis is now regarded as <1% to
2%. Accounting for this are the 2 cardinal changes in approach: more
optimal deployment by use of higher balloon dilatation pressures, and
combination antiplatelet therapy with aspirin and ticlopidine. The
marked reduction in the incidence of stent thrombosis and the
simplification of the pharmacological regimen has greatly reduced the
cost of stenting by making overnight hospital stays routine, along with
the possibility of transforming stent deployment to an outpatient
procedure in select patients.21
View this table:
[in a new window]
Table 1. Stent Randomized Thromboprophylaxis
Trials
Parallel to the improvement in pharmacological coverage to protect
against stent thrombosis, a new class of drugs was being explored for
balloon PTCA. For several years, it had been known that a final common
pathway for platelet aggregation existed. Elucidation of the
biology of this glycoprotein IIb/IIIa receptor, the most
densely expressed adhesion molecule on the platelet surface, laid
the groundwork for Coller24 to develop a
monoclonal antibody.25 The antibody was
the first agent to enter clinical trials, later followed by a peptide
and several small-molecule competitive inhibitors of this
adhesion molecule receptor. .26 27 28 29 30 31 The
Evaluation of IIb/IIIa platelet receptor antagonist 7E3
in Preventing Ischemic Complications (EPIC)
trial26 was the first "proof of concept" and
demonstrated a marked, 36% reduction in the incidence of death or MI
within 30 days of the procedure (Figure 1). This trial was conducted in high-risk
patients, eligible because of an acute coronary syndrome or
known angiographic liabilities for acute
complications.25 Either PTCA or directional
atherectomy was the revascularization procedure. Of
note, only the bolus and sustained infusion (>12 hours) but not the
bolus per se strategy was effective in reducing adverse outcomes. This
strongly suggested that the duration of IIb/IIIa blockade would play
out as an important modulator of therapeutic potential.
View this table:
[in a new window]
Table 2. Principal Features of the 6 Large Trials of IIb/IIIa
Inhibitors in Interventional
Cardiology
View larger version (16K):
[in a new window]
Figure 1. Death or nonfatal MI at 30 days for 5 trials of
IIb/IIIa inhibitors in interventional
cardiology procedures. EPIC indicates Evaluation of
IIb/IIIa platelet receptor antagonist 7E3 in Preventing
Ischemic Complications; IMPACT, Integrilin to Manage
Platelet Aggregation to prevent Coronary Thrombosis;
RESTORE, Randomized Efficacy Study of Tirofiban for Outcomes and
Restenosis; CAPTURE, Chimeric 7E3 Anti-Platelet in Unstable
angina Refractory to standard treatment; EPILOG, Evaluation of PTCA to
Improve Long-term Outcome by c7E3 GPIIb/IIIa receptor blockade; and
EPISTENT, Evaluation of Platelet IIb/IIIa Inhibitor for
Stenting. 
), with reduction of major adverse
events. Second, the field of patients was expanded to include any
routine percutaneous intervention rather than only a
high-risk inclusion criterion. Third, the bleeding complications
encountered in EPIC were related to excessive heparin and prolonged
indwelling vascular sheath time. In the trials that followed, using
lower doses of weight-adjusted heparin and reducing the time of
vascular sheaths, the initial doubling of bleeding in the patients
receiving IIb/IIIa was reduced to levels similar to those in the
control group.30
By virtue of the benefit of 2 different and distinct mechanisms,
the concept of using both arterial scaffolding and
protection from platelet thrombosis represents an
attractive clinical strategy. Supportive data for this combination in
patients receiving stents were demonstrated in subgroups of randomized
trials.32 33 As shown in Figure 2, in the Evaluation of PTCA to Improve
Long-term Outcome by c7E3 GPIIb/IIIa Receptor Blockade (EPILOG) trial,
>300 patients underwent stenting because of suboptimal balloon
angioplasty results, and the benefit for IIb/IIIa inhibition with
respect to reduction of MI and emergency
revascularization was
striking.32 Of note, the favorable effects for
combined IIb/IIIa inhibition (administered on a
prophylactic basis) and stenting had thus far been shown
only in patients with poor or suboptimal initial results. A dedicated
trial (known as Evaluation of Platelet IIb/IIIa
Inhibitor for Stenting [EPISTENT]) studied the use of
IIb/IIIa blockade with elective stenting.31 In
this trial, balloon angioplasty and abciximab was compared with either
stenting and placebo or stenting and abciximab in a total of 2399
patients. For the combination of stenting and abciximab, compared with
stenting and placebo, a striking (>50%) reduction of death and MI at
30 days and 6 months was demonstrated.31 The
balloon-abciximab arm also was significantly improved compared with
stenting and placebo.31 These findings accentuate
the need for improved antiplatelet therapy with
percutaneous intervention, whether it be with stenting
or balloon angioplasty.
View larger version (16K):
[in a new window]
Figure 2. Reduction of death, MI, or need for urgent
revascularization in prophylactic
abciximab arm of a subgroup of 322 patients in EPILOG who underwent
stenting for suboptimal results. ). Each trial that was performed in an
era in which stents were available showed a lesser need, with an
overall 24% reduction. All of these trials were performed in a
double-blind fashion. This, along with the consistent findings
across trials and the magnitude of the reduction, bespeaks a prominent
effect of improved antiplatelet coverage for achieving better
results in percutaneous intervention without stenting.
Whether indeed stenting should be avoided when possible or actually
applied more frequently is an unsettled issue.35
Notwithstanding the provisional stent debate, the interaction of
stenting and antiplatelet therapy is clearly a critical one for
future optimization of patient outcomes.
View larger version (17K):
[in a new window]
Figure 3. Need for stenting for bailout or suboptimal
results in double-blind, placebo-controlled trials of IIb/IIIa
inhibitors for interventional cardiology.
RAPPORT indicates Reo Pro and Primary PTCA Organization and Randomized
Trial; other abbreviations as in Figure 1 .
Initially, it was thought that IIb/IIIa blockade might have a
salutary effect on clinical restenosis.36
This was based on the long-term effects of the original EPIC trial, in
which there was a 24% reduction of target-vessel
revascularization at 6 months in the abciximab
bolus and infusion treatment group.36
Furthermore, this benefit was sustained even at 3 years of
follow-up.37 Theoretically, this occurred on the
basis of potent platelet inhibition before and for the first few
days after the arterial dilatation. By attenuation of the
platelet-aggregation response, less growth factor (such as
platelet-derived growth factor) and vasoactive amine (eg,
serotonin, thromboxane
A2) would be released. In addition, abciximab is
known to inhibit the vitronectin receptor, an integrin with
homology to IIb/IIIa because of the shared ß-3 subunit and >70% of
the 
-subunit. Because abciximab cross-reacts fully with the
vitronectin receptor38 39 and this is
a critical pathway for governing smooth muscle cell migration,
theoretically a dual approach of platelet and smooth muscle cell
antagonism was operational. However, when the same IIb/IIIa
inhibitor was used in the subsequent EPILOG trial, there
was no 6-month effect on target-vessel
revascularization.30 This was
also the case in both the ERASER and RAPPORT
trials,40 41 as summarized in Table 3.
View this table:
[in a new window]
Table 3. Major 6-Month End Points of the 5 Trials of IIb/IIIa
Blockade in Interventional
Cardiology , however, there was
a pronounced change in the need for target-vessel
revascularization from the initial to the more
contemporary abciximab trial. This reduction in the placebo group of
the respective trials is quite impressive and probably reflects the era
of more aggressive dilatation that has characterized
percutaneous intervention since backup stenting was
widely available in 1995. More recently, however, data from the
EPISTENT 6 month follow-up results have become available. An overall
18% reduction in the need for repeat target vessel revascularization
was demonstrated for abciximab compared with placebo in stent assigned
patients (from 10.6% to 8.7%) with >50% reduction in diabetic
patients (16.6% vs 8.1%, P=0.02). Thus, in-stent
restenosis, especially in diabetics, is favorably influenced by
abciximab.31
View larger version (20K):
[in a new window]
Figure 4. Proportion of patients requiring repeat
target-vessel revascularization in control groups
of EPIC trial (conducted in 1992; control group, n=639) vs EPILOG trial
(conducted primarily in 1995; control group, n=939) shows marked
reduction in more contemporary trial (TVR required in 22.3% in EPIC,
16.6% in EPILOG).
Since the first randomized trial of directional atherectomy versus
balloon angioplasty,42 debate has arisen as to
the significance of periprocedural non–Q-wave MI, which clearly occurs
more frequently than expected. In aggregate, a number of studies with
ample follow-up have shown that patients with a periprocedural MI
characterized by a 3-fold elevation of total creatine kinase (CK),
positive for myocardial band (MB), have a compromised long-term
prognosis.43 The patients most apt to have a
periprocedural non–Q-wave MI are those with diffuse atherosclerotic
involvement, with long lesions or saphenous vein graft target
stenoses, along with procedures that induce a deeper level of
arterial injury, particularly
atherectomy.44 45 Common to these higher-risk
subsets is the responsiveness of IIb/IIIa inhibitors for
the reduction of periprocedural MI. Although the predominant
therapeutic effect of the IIb/IIIa inhibitors is in large
MIs, Q-wave or non–Q-wave, there is clear-cut reduction of
periprocedural MIs in the range of 3- to 5-fold CK
elevation.26 27 28 29 30 31 Accordingly, using this drug
class as a mechanistic probe, we know that at least some of these
events are attributable to platelet aggregation, whether it be the
site of arterial dilatation or in the microcirculation. The
weight of evidence suggests that the chief hazard of periprocedural MI
is increased death rate during extended
follow-up,43 which appears to be directly related
to the size of the MI. This may explain the favorable impact of
IIb/IIIa blockade on mortality during 3-year follow-up of a cohort of
patients with acute coronary syndromes in the initial EPIC
trial.37 Nevertheless, periprocedural MIs still
occur despite full doses of IIb/IIIa inhibitors, so that
other factors, such as side-branch closure, atherosclerotic emboli, and
overriding intimal disruption, play a role in some patients.
One of the most exciting frontiers in
cardiovascular medicine is the ability to disaggregate
platelets in the setting of acute MI.46 The
underlying pathophysiology has been further elucidated, with the white
clot central to the response to plaque fissure or erosion and the
attendant exposure of subendothelial matrix. Until
recently, there was no tool in our therapeutic armamentarium to
dissolve the platelet clot—that is, to achieve "platelet
lysis." Indeed, the prothrombotic effects of plasminogen
activators led to failure of angioplasty in the early
randomized trials.46 47 48 49 This has been completely
turned around by exploiting our newfound ability to dissolve white clot
in the acute setting, with or without fibrinolytic therapy. Use of
IIb/IIIa inhibition alone has resulted in infarct vessel patency rates
approximating the use of streptokinase
therapy.50 ). Neumann and
colleagues,51 from Munich, Germany, reported a
trial of 200 patients undergoing primary stenting for acute MI with a
similar reduction of adverse outcomes with abciximab compared with
conventional pharmacological therapy. Recent data from dose-finding
trials in the TIMI-14 and GUSTO-4 trials indicate that half-dosing of
tissue plasminogen activator (alteplase) or
recombinant tissue plasminogen activator
(reteplase), respectively, given in combination with full-dose
abciximab, can achieve infarct vessel patency in >90% of patients by
60 to 90 minutes into therapy.52 53 In the
related trials of acute coronary syndromes, the patients
receiving IIb/IIIa blockers on presentation who had
subsequent percutaneous coronary intervention
during drug infusion derived the most benefit of all of the subgroups
analyzed (Table 4).54 55 56
View larger version (15K):
[in a new window]
Figure 5. RAPPORT trial event rates of death, MI, and urgent
revascularization for abciximab vs placebo.
View this table:
[in a new window]
Table 4. Use of Percutaneous Coronary Revascularization While
on IIb/IIIa Blockade Therapy in 3 Trials of Empirical Therapy in Acute
Coronary Syndromes ).74 75 76 77 78 79 80 81 82 83 84 The
results of a single-center clinical trial performed in Rotterdam that
examined the effectiveness of antisense compound directed against
messenger RNA for the protein product of the c-myc gene
will soon be reported. This will be the first human trial to use an
antisense DNA strategy for the prevention of restenosis.
View this table:
[in a new window]
Table 5. In Vivo Antisense
Studies 
50% of all coronary angioplasty procedures. This
"stentomania" is driven both by the gratifying short-term results
seen by the interventionalist using these devices and by results
extrapolated from observational and randomized trials. The "bigger is
better" dogma first proposed by Kuntz and his
colleagues85 has been adopted by many operators,
who feel that the greater the acute gain in lumen diameter, the smaller
the chances of short- and long-term failure.86
Inherent in this concept is that the stent is merely a means to an end,
and therefore, irrespective of the particular stent or other dilating
or debulking device used, success of the procedure is determined solely
by the acute results. Whether differences between various stent designs
and materials are sufficient to impact their clinical effects remains
to be demonstrated in controlled "stent versus stent" trials. The
positive results of the few completed randomized trials have been
enthusiastically applied to almost every other patient and lesion
subcategory. The unequivocal indications for stenting are currently
only those few that have been supported by observational and randomized
trials and are limited to very few stent types. The definitive evidence
for the use of stents for several specific clinical indications is
still lacking. As results of the many ongoing randomized stent trials
become available, the indications for stenting will have to be adapted
accordingly. Currently, there is solid evidence from observational
studies and randomized trials to support the use of coronary
stents for the following indications: (1) treatment of abrupt or
threatened vessel closure during angioplasty, (2) primary reduction in
restenosis in de novo focal lesions in vessels >3.0 mm in
diameter, (3) focal lesions in saphenous vein grafts, (4) totally
occluded vessels, and (5) treatment of acute MI.
With the current and future plethora of intracoronary
stenting devices, the obvious question that must be answered is whether
the choice of a particular stent makes a difference with respect to
clinical outcome. Two lines of evidence, 1 experimental and the other
clinical, suggest that a difference does indeed exist between different
stents.87 88 89 90 In an animal model, it has been
suggested that stent surface material and geometric configuration may
be more important than operator-dependent variables in determining
the degree of neointimal hyperplasia and
thrombosis.87 Design characteristics such as hoop
strength91 and metallic surface
area92 have been shown to influence
neointimal formation in experimental models. The metal
composition and characteristics of the stent surface may also be
important for the performance of the implanted
stent.93 94 95 96 97 98 99 100 101 102 ) outcomes were demonstrated for the
Multi-Link and the Palmaz-Schatz stents. Likewise, clinical and
angiographic (Table 6) equivalence between the Palmaz-Schatz stent and
the Micro Stent II was shown in the Study of AVE-Micro Stent Ability to
Limit Restenosis (SMART) trial, in which 613 patients with
focal de novo or restenotic native coronary lesions
were randomized. Equivalence between the GR II and the Palmaz-Schatz
stent was not demonstrated in the GR II trial. Differences at 6-month
follow-up were seen in the 6-month clinical outcomes of target-vessel
revascularization and target-vessel failure
(composite of death, MI, and target-vessel
revascularization) (Figure 6). Significant differences were also
seen in the 6-month follow-up angiographic parameters of
minimal luminal diameter, percent diameter stenosis, and binary
restenosis rate (Table 6). The reasons for these differences
are not clear but may be due in part to a higher acute residual
stenosis after stent implantation, a higher loss index in the
GR II–treated vessels (0.76 versus 0.57; P=0.007), or as a
result of GR II stent undersizing or longer stent length in the GR II
group. With the large number of stents currently available, it is
questionable what practical purpose will be served by comparing all the
available stents with the "standard" (Palmaz-Schatz) or with each
other. Also, is similar performance at 6-month follow-up
angiography in a select patient population adequate to indicate
equivalency? Other issues that must be considered when evaluating a
stent, other than equivalency to an arbitrary standard, are ease of
use, versatility, and cost.
View this table:
[in a new window]
Table 6. Angiographic Follow-Up of Stent Versus Stent
Trials
View larger version (13K):
[in a new window]
Figure 6. Clinical outcome of GR II trial comparing GR II
stent (n=364) to Palmaz-Schatz stent (n=364) showing frequency of death
and subacute thrombosis (SAT) at 30 days and frequency of death,
target-vessel revascularization (TVR), and
target-vessel failure (TVF) (death, TVI, MI) at 9 months. Differences
can be seen in frequency of both TVF and TVR between Palmaz-Schatz
stent and GR II stent.
With recent improvements in deployment techniques for
intracoronary stents and increased operator experience, lesions
previously considered not amenable to percutaneous
treatment are now being treated with intracoronary stent
implantation. Industry has responded to the demand with the
production of myriad customized stents for very particular
applications.
Several new stent designs are available that are constructed
specifically for use in bifurcation lesions. The Jostent B (Jomed
International AB) is 1 such device. The configuration and size of the
cells at 1 end of the Jostent B is similar to the first-generation
Jostent M stent, with the cells connected with V-shaped bridges. At the
other end of the stent, the cells are larger and connected with
straight bridges (Figure 7A). On full
expansion, the larger cells have a diameter of 3.5 mm, allowing
easy access to the other arm of the vessel bifurcation.
View larger version (139K):
[in a new window]
Figure 7. A, Photograph of Jostent B showing larger cells at
1 extreme of stent. B, Photograph of Bard Bifurcation Stent, a true
bifurcated device. C, Photograph of Devon Side-Arm Stent showing
region with omitted struts. D, Photograph of Devon Ostial
Stent showing reinforced struts at 1 extreme of stent. E, Photograph of
end of a Jostent Coronary Stent Graft showing expandable PTFE
material sandwiched between 2 thin-strut metal stents. ). The
main body of the stent is a single coil, through which the 2 balloons
pass. The balloons diverge at the crux of the Y to pass separately
through the 2 arms. Implantation into animals is currently under way,
and the first human implants are being planned.
Stents specifically designed for use in the treatment of lesions
at the site of significant side branches are also available. SciMed
Life Systems supplies the NIRSide, which has the same basic design as
the standard family of NIR stents; however, the cells in the center of
the stent are larger than those at the ends. When properly positioned,
the number of obstructing struts over the entrance to the side branch
is minimized and side-branch access is facilitated. The design concept
of the Jostent S (Jomed International AB) is similar to that of the
NIRSide stent (Figure 7C).
Devon Medical produces a stent designed exclusively for use in
ostial lesions. The base design for the stent is the Pura-Vario A. At 1
end of the stent, however, the terminal row of cells is slightly longer
and the struts slightly thicker (Figure 7D). This not only increases
the radial force of this portion of the stent but also increases its
radiopacity, which facilitates precise positioning.
The Jostent Coronary Stent Graft (Jomed International AB)
is a unique integration of graft material into a coronary
stent. This device was constructed with a sandwich technique whereby an
ultrathin layer of expandable PTFE, specially developed for integration
into a stent graft system, is placed between 2 stents with reduced
strut thickness (Figure 7E). The Stent Graft is also offered coated
with the Corline Heparin Surface, which has the potential to reduce the
risk of thrombus formation after stent implantation.
The use of stents as platform for the delivery of radiation to the
vessel wall has been receiving considerable
attention.113 114 Stent-bound radioactive sources
can deliver effective doses of radioactivity to all levels of the
vessel wall. It is also believed that radioactive stents can act by
culling the smooth muscle cell population as these cells pass through
the "electron fence" at the plane of the stent
wires.115 
- and
ß-radiation from the radionuclides 55,56,57Co,
52Mg, and 55Fe. Stents with
activities of 3.9, 17.5, and 35 mCi were tested. At 4 weeks, exposure
to the 2 higher dose levels resulted in a significant reduction in
neointimal formation, and all treated animals exhibited a
significant reduction in proliferating cell nuclear antigen–positive
cell and smooth muscle cell counts. Vascular
reendothelialization occurred despite prolonged
irradiation; however, the time to complete endothelial
cell coverage was delayed in a dose-dependent manner.
The list of materials used to coat metal stents in an attempt to
reduce their inherent thrombogenicity and decrease the incidence of
in-stent restenosis is long and ever
increasing124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 (Table 7). A few, however, deserve special
attention. Most coatings tested are placed mainly to provide a
biologically inert barrier between the stent surface and the
circulating blood. Commercially available are the gold-coated InFlow
and NIR stents and the silicon carbide–coated Tensum and Tenax stents.
In contrast to these, immobilized-heparin surface coatings
have been studied as a means of providing a biologically active
exterior that interacts with the circulating blood. Many techniques
have been applied to attach heparin to synthetic surfaces; however, the
description of a process for end-point attachment of heparin to
polymer-coated surfaces that preserves the activity of the antithrombin
binding site made the production of heparin-coated stents
feasible.143 Heparin-coated stents were shown to
be effective in reducing thrombosis in rabbit peripheral
vessels135 and in porcine coronary
arteries.136 137 Three heparin-coated stents are
currently available for clinical use: the Cordis/Johnson & Johnson
heparin-coated Palmaz-Schatz stent, on which heparin is end-linked to
the stent surface with a patented Carmeda coating technology; the
Wiktor heparin-coated stent (Hepamed coating)144;
and the Jostent (Corline heparin coating), on which heparin is randomly
attached. Random covalent binding of heparin to the stent surface
results in variable exposure of antithrombin binding sites, whereas
high-activity end-point attachment, as on the Cordis/Johnson & Johnson
product, ensures that all of the anti–thrombin III binding sites
remain functionally intact. Animal studies using the high-activity
heparin-coated stents have shown that up to 80% of the antithrombin
III binding activity is lost 4 weeks after stent
implantation.145 Nevertheless, reduction in rates
of stent thrombosis in animal studies led to the evaluation of the
high-activity end-point–attached heparin-coated stents in the
BENESTENT II pilot study146 and the BENESTENT II
randomized trial.147 In the 616 patients
receiving a heparinized stent in these studies, there was only 1
episode of subacute thrombosis (incidence <0.2%).
View this table:
[in a new window]
Table 7. Coating Material Considered for Use With Metal
Intracoronary Stents
Despite the phenomenal pace of stent design technology, the
incidence of stent restenosis remains unacceptably high. To
address this problem, intense investigation into new stent coatings
continues, with new biocompatible and drug-eluting polymers being
developed for application to the metal stent scaffold. As novel
biocompatible drug delivery stent coatings are developed,
pharmacological compounds that failed to prevent restenosis
when administered systemically and those with significant systemic
toxicity are now being reexamined. The new drug-eluting stents also
provide a unique platform for the administration of proteinaceous
compounds, antisense oligonucleotides, and DNA that
cannot be given systemically. In addition, metal stents provide a
convenient means for the application of brachytherapy by use of
ionizing radiation. Although initial clinical results seem promising,
issues of correct dosimetry and the long-term effects of cytotoxic
therapies on the arterial wall have arisen. Despite recent
significant strides in stent design technology, it is apparent that
much has yet to be learned and further advances are necessary to
improve the long-term outcome of patients treated with intravascular
stents. , these salutary effects were
also associated with less need for repeat hospitalization and less
antianginal medication. For example, in the study by Allen et
al,187 the need for cardiac hospitalization at 3
months was reduced from 43% in medically managed patients to 20% in
the group treated with TMR. Thus, it appears that TMR is effective for
reducing angina, but the mechanism is not fully elucidated. Because the
holes occlude fairly quickly, over just a few weeks in postmortem
studies and experimental models, the most likely explanation for the
benefit is angiogenesis.168 169 170 188 However,
neuronal dysfunction induced by the laser has been raised as a
plausible explanation for some of the antianginal effect based on
elegant experimental modeling that focused on reflexes to systemic
hypotension with bradykinin applied directly to the epicardium after
TMR.189 Recent mechanistic studies have verified
improvement in dobutamine-induced angina threshold,
myocardial blood flow by PET, and even on indices of
ventricular repolarization.190 191 192
With regard to safety, there were some initial concerns regarding a
high operative mortality in these series (in the 8% to 10% range),
but more recent experience suggests that this may have been
attributable to patient selection or an obligatory operator learning
curve.
View larger version (15K):
[in a new window]
Figure 8. Incidence of class I or II angina among patients
randomized in 2 trials of transmyocardial
revascularization (TMR) vs medical therapy (Med)
(References 74 and 75). All patients had class 3 or 4 angina at time of
enrollment.
The field of interventional cardiology is poised
to go through significant further development. By the start of the
2000s, improved platelet coverage will be routine in all patients
undergoing procedures, and oral therapies for more extended periods
during follow-up will be more thoroughly investigated. It would not be
at all surprising if angiogenesis therapy, be it by
intracoronary, intramyocardial (at the time of surgery), or
transmyocardial laser revascularization, became an
important adjunctive means of achieving improved myocardial blood flow.
Restenosis therapies will undoubtedly blossom beyond the
initial application of radiation for in-stent restenosis,
ideally taking advantage of the increased understanding that we now
have of genetic control of cell proliferation and matrix
production. Although stents are already being used in >60% of
patients at the time of interventional procedures, their use will also
increase further with improved designs, coatings, and the newfound
ability to use the stent as an effective drug delivery station.