Dr. Agah is a Fellow in Interventional Cardiology at the
Cleveland Clinic Foundation, Cleveland, OH.
The last decade has seen tremendous growth in basic research
attempting to target disease at the deoxyribonucleic acid (DNA)
level. At present, there are more than 500 trials in gene therapy.
Approximately 50 of these are cardiovascular studies,
1
which target treatment for coronary artery disease, peripheral
vascular disease, and restenosis. Most current gene therapies
require invasive measures to deliver the treatment vector locally.
Hence, once these investigational tools become a clinical reality,
the invasive/interventional cardiologist would play a central role
in the delivery of these treatments.
However, the early promise offered by gene therapy has given way
to cautious optimism as the field has matured. The past year best
represents a watershed in this process as the Food and Drug
Administration suspended all ongoing gene therapy trials involving
adenovirus, following the death of an 18-year-old subject in a
phase I trial targeting therapy for ornithine transcarbamylase
deficiency. This suspension included several trials in
cardiovascular medicine targeting treatment for coronary and
peripheral vascular diseases using angiogenic factors. Following
extensive review of present protocols by panels of experts, these
trials have recently resumed with stricter guidelines. The purpose
of this review is to assess the present status of cardiovascular
gene therapy, following a short review of the fundamentals of
gene-based therapy.
Fundamentals of gene therapy
The basis of gene therapy lies in altering the pathophysiology
of the disease process by introduction of DNA for a specific gene
into cells.
2
There are several prerequisites for this process: 1) understanding
the molecular mechanisms involved in the disease process and areas
for potential targeted therapy; 2) isolation of the gene of
interest for targeted therapy (cloning of the gene); 3)
availability of vectors to introduce the gene of interest into the
targeted cells; and 4) expression of the therapeutic gene in the
host cells in a manner to produce the desired effect with minimal
toxicity to the organism.
The success of the latter two points relies in development of
"ideal" gene therapy vectors. Although there has been significant
progress in the field in the last decade, the development of such
vectors remains the "Achilles' heel" of the field of gene therapy.
3
These vectors fall into two broad categories: the viral and
nonviral vectors. Although the nonviral vectors can be generated in
large amounts and are generally safer, they remain inefficient in
the transfer of genetic material.
At present, most gene therapy trials rely on viral vectors.
4
The viral vectors are generated on the premise of cloning the gene
of interest in the virus and using the natural ability of the virus
in infecting host cells to deliver the (therapeutic) gene into the
cells. To abrogate the cytopathic effect, the virus is made
replication-deficient by removing the genetic component necessary
for viral replication (figure 1). These altered viruses are
replication-incompetent, making them safe for human use. The
recombinant viruses can be grown to high titer for infection of a
large number of cells. Once grown to high titer in culture and
purified, they can be introduced in vivo to infect the targeted
cells and express the gene of interest.
Restenosis
Restenosis appears to be an "ideal disease" for gene-targeted
therapy in cardiovascular medicine.
5
The tissue injury response to angioplasty is one of smooth muscle
cell migration-proliferation and constrictive remodeling. While
both processes take place with balloon angioplasty, only the former
has a significant role in in-stent restenosis. With wide
application of stents and late treatment of balloon angioplasty
"failures" with placement of stents, the restenotic process in
percutaneous coronary intervention has become mainly one of
neointimal hyperplasia-smooth muscle proliferation and migration.
With expanded knowledge of factors involved in cell cycle division
and ligands that can activate these factors (cytokines, growth
factors, transcription factors) strategies for inhibiting cell
division by blocking these factors have emerged (Table 1). The
successful approaches have included delivery of anti-sense
oligonucleotides, naked DNA, and recombinant adenovirus introducing
the recombinant gene able to inhibit cell division. When used in
vivo, these vectors have been able to prevent restenosis in
multiple animal models, including the rat carotid artery, the
rabbit iliac artery, and the porcine femoral and coronary arteries.
6-8
Despite their success in inhibiting restenosis in animal models,
early experience in human studies has been disappointing. The
reasons primarily appear to be inefficient local delivery of
vectors via conventional interventional techniques and low
transfection/infection efficiency of the vectors. This early
failure, along with emerging alternative treatment for restenosis
(including brachytherapy and coated stents), has dampened the
initial flood of gene therapy trials for restenosis, with only one
ongoing trial still in progress in the United States.
Heart failure
In contrast to declining mortality from coronary disease, deaths
from heart failure have doubled in the last 2 decades. As most of
these patients are elderly and are not candidates for cardiac
transplantation, medical treatment remains the only option
available to these patients. As such, these patients are ideal
candidates for percutaneous gene therapy approaches.
There are several obstacles that need to be addressed in any
approach for gene-based therapies in heart failure, however: 1)
widespread, if not uniform, transduction of the cardiac cells by
the vectors and the delivery technique; 2) long-term
expression/effect of the delivered gene and/or ability for
readministration; and 3) targeted therapy for the underlying
pathophysiologic mechanism of heart failure, i.e., systolic versus
diastolic dysfunction. We shall address the progress in each of
these areas below.
9
The formidable task of delivering the gene therapy vector to the
billions of cardiac cells is potentially remedied by the rich
vascular supply to this organ via the coronary arteries. Earlier
reports that coronary injection of adenoviral vectors can transduce
10% to 30% of cardiac myocytes have given way to modified
approaches where now it is possible to transduce greater than 50%
of all cardiac cells. These techniques have included transient
cross-clamping of the aorta and pulmonary artery after injection of
the viral vector in the left ventricular chamber. A less robust but
easier approach involves cross-clamping only the aorta during the
infusion. Although the percutaneous approach for such techniques is
not yet established in humans, it may be clinically tolerated in
the same way that the aortic occlusion during aortic valvuloplasty
is well tolerated by patients for a short duration of 10 to 20
seconds.
In contrast to gene therapy approaches in restenosis, in which
transient expression of the therapeutic gene may be sufficient for
the treatment effect, most protocols for therapies in heart failure
require long-term expression of the therapeutic gene. Although
recombinant adenovirus has been a natural choice for restenosis
studies, the early generation adenoviral vectors are inadequate for
treatments of heart failure due to the transient nature of gene
expression. Furthermore, since adenoviral gene expression is
eliminated primarily via an immune response to adenoviral genes,
such global cardiac immune response has the potential for
myocarditis. As such, gene therapy protocols in heart failure had
to await the development of a vector with minimal induction of host
immune response upon delivery, ability for long-term expression of
the transgene, and sustained and efficient transduction of the
cardiac myocytes. These vectors have come of age in the last
several years. One such vector has been the third-generation
adenoviral vector, from which most of the viral "backbone" has been
eliminated as well, and hence the accompanying immune response.
Such modification has led to long-term expression of the
transgene.
Another vector with potential application for cardiac gene
therapy is the adeno-associated virus; this vector has the ability
to transduce cardiac myocytes with similar efficiency as
adenovirus, but does not elicit an immune response that eliminates
the transduced cells. As such, they can have long-term expression
in cardiac tissue after single administration. However, they are
difficult to grow to high titers in tissue culture.
The study of cardiac myocytes in failing heart has revealed
multiple molecular abnormalities leading to depressed contractile
function of these cells. Broadly, two signaling pathways involved
in contractility have been targeted so far. One pathway involves
intracellular calcium regulation. There is intracellular release of
calcium in the myocytes with electrical excitation leading to
myofibril contraction. During myofibril relaxation, calcium is
extruded extracellularly mainly by the sarcoplasmic reticulum
calcium adenosinetriphosphatase (ATPase) pump (SERC2a). In turn,
the activity of SERC2a pump is regulated by phospholamban. Studies
of single myocytes from failing hearts have revealed elevated
end-diastolic calcium levels in these cells associated with reduced
systolic force, elevated diastolic force, and abnormal relaxation.
These abnormalities in calcium regulation are associated with
decreased ratio of SERC2a to phospholamban. Several in-vitro and
in-vivo studies have shown that restoring the effective
SERC2a/phospholamban ratio can restore myocyte contractility and
systolic function in animal models of heart failure.
10
These studies have paved the way for targeting SERC2a and
phospholamban to failing human heart cells.
Another signaling pathway for targeted therapy, with a better
clinical correlate for therapeutic efficacy, is the beta
(ß)-adrenergic signaling. It is well known that in heart failure
there is down regulation of the myocardial ß-adrenergic receptor
(ß-AR) along with upregulation of an inhibitor of receptor
signaling ß-AR kinase-1 (ßarK-1). One of the clinical benefits of
administering oral ß-blockers in heart failure patients is thought
to be secondary to restoring this signaling mechanism. Several
studies have shown that adenoviral delivery of ßarK-1 can rescue
the heart failure phenotypes in various animal models.
11
The concept of apoptosis as the final common molecular pathway
in the diverse spectrum of heart failure has sparked interest in
local delivery of various anti-apoptotic factors to prevent
progression of cardiac cell death. However, data demonstrating the
benefit of such "rescue" in relevant animal models is not available
at this time.
Angiogenesis
Treatment of ischemia by increasing the number of small vessels
in the ischemic tissue--angiogenesis--has been the focal point of
research for patients not candidates for traditional means of
revascularization. No area in cardiovascular gene therapy has
generated as much interest and public enthusiasm as angiogenesis.
This is particularly remarkable considering the limited amount of
data available to demonstrate a clear clinical benefit of such an
approach.
12
Because of the absence of published placebo-controlled, randomized
clinical trials (although several are in progress), we shall limit
this review to studies of angiogenesis in animal models. As a
whole, most studies have concentrated on two families of angiogenic
factors: vascular endothelial growth factor (VEGF) and fibroblast
growth factors (FGF). These factors have been delivered as
recombinant protein, plasmid DNA, and adenovirus encoding the cDNA
of the growth factor. The routes of administration have also
varied, but typically can be categorized as direct intramyocardial
injection versus intracoronary injection.
In a study by Rosengart et al
13
using an ischemic porcine model, these authors were able to show
that 4 weeks after injection of adenovirus (Ad.) carrying the VEGF
gene, Ad.VEGF-121, both regional function (echocardiography) and
perfusion (radionuclide imaging) were improved in the
Ad.VEGF-121-treated animals when compared with control. Similar
results that related improvement in regional blood flow (using
microspheres) were published by Isner et al, with injection of
plasmid DNA (ph.VEGF-165).
14
Hammond et al,
15
using the same porcine model, delivered Ad.FGF-5 to the ischemic
area using an intracoronary approach. Two weeks after gene
transfer, there was improvement in stress-induced ischemia
(assessed by echo) in the Ad.FGF-5-treated animals compared with
controls. The results of these animal studies are supported by
uncontrolled observational reports where injection of angiogenic
factors in the ischemic myocardium results in subjective
improvements in patients' symptoms. Currently there are multiple
ongoing phase 2 and phase 3 clinical trials evaluating the efficacy
of both FGF and VEGF in improving ischemia in patients not
candidates for traditional revascularization. The results of these
trials are awaited with great anticipation, as they may establish
the first gene-based therapy in clinical cardiology.
Graft failure
Vein graft failure after coronary artery bypass graft surgery
and peripheral bypass surgery may be the ideal clinical setting for
gene therapy applications. The excision of the graft from the host
allows the tissue to be manipulated "ex vivo" before it is
re-implanted in the subject.
16
As such, uniform and controlled transgene transduction can be made
before re-implantation. As we gain more knowledge in the
pathophysiology of graft failure and its distinct nature from
atherosclerosis, we will be better able to target therapy to this
clinical problem. Preliminary studies in animal models targeting
antiproliferative and antimatrix agents using adenoviral and
retroviral vectors has met with initial success in limiting graft
restenosis and subsequent failure. Noteworthy in these approaches
is the result of the PREVENT study.
17
This was a single-center, phase 1 trial to assess the safety of
using an anti-sense oligonucleotide decoy against the cell cycle
regulator gene, E2F, to block neointima growth and prevent graft
failure of peripheral arterial bypass grafts in high-risk patients.
Although the study was not designed to assess clinical outcome, at
1 year patients in the treatment group had a 29% rate of graft
failure compared with 69% in the control group. These encouraging
early results await validation by large-scale randomized
trials.
Atherosclerosis and hypertension
No direct gene therapy approach to atherosclerosis exists today
due to the multifactorial, multistage, and insidious nature of this
disease. However, directed therapies against hypercholesterolemia,
a major risk factor for development of atherosclerosis, are in
progress with encouraging early results in animal models
18,19
and a small phase 1 trial in humans.
20
Similar approaches exist for the treatment of hypertension with
early success in animal models.
21
However, progress in both these areas is slow due to the existence
of effective medical therapy for both conditions.
Conclusion
With experience from the past decade, the field of gene therapy
is making steady progress toward becoming a clinical reality. With
active research and ongoing trials in heart failure, restenosis,
and angiogenesis, the next decade may see the first successful
application of this novel therapy in a broad spectrum of
patients.