Cardiovascular Gene Therapy: Past, Present, and Future

Abstract

Cardiovascular diseases remain a large global health problem. Although several conventional small-molecule treatments are available for common cardiovascular problems, gene therapy is a potential treatment option for acquired and inherited cardiovascular diseases that remain with unmet clinical needs. Among potential targets for gene therapy are severe cardiac and peripheral ischemia, heart failure, vein graft failure, and some forms of dyslipidemias. The first approved gene therapy in the Western world was indicated for lipoprotein lipase deficiency, which causes high plasma triglyceride levels. With improved gene delivery methods and more efficient vectors, together with interventional transgene strategies aligned for a better understanding of the pathophysiology of these diseases, new approaches are currently tested for safety and efficacy in clinical trials. In this article, we integrate a historical perspective with recent advances that will likely affect clinical development in this research area.

Ylä-Herttuala and Baker summarize recent advances and the current status of cardiovascular gene therapy with special emphasis on new clinical initiatives and lessons learned from previous clinical trials, including a discussion of the selection of patient populations, endpoints, and best available treatment strategies.

Main Text

The potential of cardiovascular gene therapy became evident in the late 1980s, when it was shown that direct intra-arterial gene transfer was possible with endovascular catheter techniques.¹ At the same time, hyperlipidemias became a target for gene therapy,² and conditions like in-stent restenosis, vein graft stenosis, heart failure, arrhythmias, refractory angina, and peripheral vascular disease were recognized as potential targets for gene therapy. In the clinics, the pioneering work of Isner et al.³ used plasmid gene transfer to treat severe peripheral vascular disease, and adenoviral vectors were, for the first time, used for local endovascular catheter-mediated gene therapy in humans.⁴

In spite of great enthusiasm and positive preclinical results, clinical translation of cardiovascular gene therapy has not yet been very successful. There are many factors that might contribute to this negativity, including insufficient gene delivery to the target site, undermining the potential of transgenes; too short a duration of transgene expression from particular transduction strategies; insufficient knowledge about underlying pathophysiological mechanisms, leading to a retrospectively poor strategy; and, often, underpowered clinical trial design. Several prominent trials that have produced negative results and failures have sent investigators back to the laboratory to re-evaluate and design better vectors and gene therapy approaches for cardiovascular diseases.5, 6 This bench-bedside iteration has been fruitful in other gene therapy arenas but has so far failed to have a significant impact in the cardiovascular space.

In spite of these difficulties, during the last 2–3 years, significant clinical and conceptual progress has been made in cardiovascular gene therapy. The first gene drug approved in the Western world, Glybera, is indicated for the treatment of severe lipoprotein lipase deficiency. Even though this condition is ultra-rare, it was an important milestone for the entire field of gene therapy.⁷ Additionally, gene delivery techniques have been significantly improved, particularly with respect to catheter-based approaches, and targeting of powerful new therapeutic genes to the myocardium has recently produced promising results.⁸ Thus, a new generation of cardiovascular clinical trials is primed to evaluate the potential of gene therapy in carefully selected patient populations. This wave of trials must include carefully documented gene transfer effects and objectively measurable changes in parameters like blood flow, metabolic activity, and cardiac function⁹ because the impact of these results cannot be underestimated. Potential targets for cardiovascular gene therapy are presented in Figure 1.

Figure 1.

Potential Targets for Cardiovascular Gene Therapy

Coronary Heart Disease

Coronary heart disease is caused by insufficient blood flow to the myocardium because of atherosclerotic stenoses in the main coronary arteries. It is well known that some patients can develop collateral arteries that rescue the myocardium in spite of the stenoses. Therefore, it is logical that this condition could be treated by providing new blood flow to the ischemic muscle.¹⁰ Coronary bypass surgery and angioplasty techniques are excellent examples of this treatment principle. However, these approaches are not possible for all patients, and there is an increasing group of severe angina pectoris patients, so-called refractory angina, who cannot be treated with these conventional approaches.¹¹

Because the basic mechanisms of angiogenesis and blood vessel formation are already well known, therapeutic vascular growth, which tries to grow new blood vessels in the ischemic myocardium, is a potential new treatment option. Several growth factors have been used to achieve this goal in the past, but, until recently, no clinically significant results have been obtained in phase II/III clinical trials.⁶ The most promising candidates for therapeutic vascular growth seem to be members of the vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF) families, hepatocyte growth factor (HGF), and gene therapy approaches combined with cell therapy.12, 13

During the last decade, several randomized controlled trials have tested either naked plasmids or adenoviral vectors for the treatment of severe coronary heart disease. Trials like EuroinjectOne,¹⁴ KAT,¹⁵ REVASC,¹⁶ NOTHERN,¹⁷ NOVA,¹⁸ VEGF-Neupogen,¹⁹ and GENASIS²⁰ have tested the intramyocardial delivery of VEGFs (Table 1). Gene transfer has also been used during bypass surgery and via minithoracotomy using epicardial injections.21, 22 Importantly, safety in these trials has been very good, even after 10-year follow-up.23, 24, 25 At the moment, there are five angiogenic gene therapy trials that are either ongoing or have recently reported final results (Table 1). Two of these trials are testing a novel VEGF-D^dNdC in refractory angina patients using percutaneous endocardial delivery of adenoviral vectors.⁸ Targeting to the ischemic hibernating myocardium is done using a combination of electroanatomical mapping and ¹⁵O-H₂O-positron emission tomography (PET), which makes it possible to treat areas that suffer from stress-induced ischemia.⁹ Crucially, absolute myocardial bloodflow can be measured with ¹⁵O-H₂O-PET and correlated to potential clinical benefits. Because VEGF-D^dNdC is both an angiogenic and lymphangiogenic growth factor, it represents a new strategy (i.e., therapeutic vascular growth) to treat refractory angina in addition to angiogenic effects: stimulation of the lymphatic circulation will help to reduce myocardial edema, which is a common side effect after proangiogenic therapies.14, 15, 16, 17, 18 VEGF-D^dNdC can also potentially induce regenerative effects by inducing stem and progenitor cells in the treated muscles.²⁶ A new trial based on intramyocardial injections via thoracotomy has also been planned. This study is based on an adenovirus expressing three major vascular endothelial growth factor-A (VEGF-A) isoforms, which should lead to a better angiogenic response by generating more natural growth factor gradients between ischemic and normoxic parts of the myocardium.²⁷

Table 1.

Ongoing or Recent Trials in CAD and PAD

Trial	Disease	Vector	Therapeutic Agent	Delivery	Study Design	n	Primary Endpoint	Main Result	Reference
KAT301	CAD	Ad	VEGF-DdNdC	percutaneous i.my. NOGA/PET guided injections	phase I, RCT	30	safety, perfusion reserve in ischemic area (PET)	positive, increased perfusion reserve in treated areas at 1 year	8
ReGenHeart	CAD	Ad	VEGF-DdNdC	percutaneous i.my. NOGA/PET guided injections	phase II, RCT	180	6-min walking test, perfusion reserve in ischemic area (PET/SPECT)	NA	NCT03039751
ASPIRE	CAD	Ad	FGF-4	percutaneous i.c. injections	phase III, open-label	100	reversible perfusion defect (SPECT)	NA	29
AWARE	CAD	Ad	FGF-4	percutaneous i.c. injections	phase III, RCT	300	ischemic ECG changes on ETT	NA	NCT00438867
VEGF-A116A	CAD	Ad	VEGF-A116A	thoracotomy i.my. injections	phase I/II, open-label	41	time to 1 mm ST depression on ETT	NA	NCT01757223
HGF-X7	CAD	Ad	HGF	percutaneous i.my. injections	phase I, open-label, no controls	12	safety	NA	30
MULTIGENE ANGIO	PAD	RV	VEGF-A165, Ang1 with cell therapy	percutaneous i.a. injections	phase I, open-label, no controls	23	safety, amputation-free survival	positive, amputation-free survival 72% at 1 year	51
JVS-100	PAD	Pl	SDF-1	i.m. injection	phase II	48	safety, amputation rate	NA	NCT01410331
HGF-X7 (VM202)	PAD	Pl	HGF	i.m. injections	phase II	50	visual analog scale	NA	NCT01064440
HGF-X7 (NL003)	PAD	Pl	HGF	i.m. injections	phase II, RCT	200	ulcer area, visual analog scale	NA	NCT01548378
KAT-PAD101	PAD	Ad	VEGF-DdNdC	i.m. injections	phase I, RCT	30	safety, perfusion in treated area (PET)	NA	EudraCT 001019-22
Neovasculgen	PAD	Pl	VEGF-A165	i.m. injections	phase II/III, RCT	100	pain-free walking distance	positive, pain-free walking distance increased 167% at 1 year	43
Previous Trials in CAD and PAD
KAT	CAD	Ad, Pl	VEGF-A165	percutaneous i.c. injection at angioplasty site	phase II, RCT	103	improved myocardial perfusion (SPECT)	positive, increased perfusion at 6 months (only adenovirus)	15
REVASC	CAD	Ad	VEGF-A121	thoracotomy i.my. injections	phase II, RCT	67	time to 1-min ST depression on ETT	positive	16
VEGF thoracotomy	CAD	Ad	VEGF-A121	thoracotomy i.my. injections	phase I, open-label, no controls	21	safety, angina classification	positive, improvement in angina classification at day 30	21
EuroinjectOne	CAD	Pl	VEGF-A165	percutaneous i.my. NOGA guided injections	phase II, RCT	74	improved myocardial perfusion (SPECT)	negative	14
Genasis	CAD	Pl	VEGF-2	percutaneous i.my. injections	phase III, RCT	295	ETT	negative	20
NORTHERN	CAD	Pl	VEGF-A165	percutaneous i.my. NOGA guided injections	phase II, RCT	93	change in myocardial perfusion (SPECT)	negative	17
NOVA	CAD	Ad	VEGF-A121	percutaneous i.my. NOGA guided injections	phase I/II, RCT	17	ETT	negative	18
VEGF-Neupogen trial	CAD	Pl	VEGF-A165, hGCSF	percutaneous i.my. NOGA guided injections and systemic hGCSF administration	phase II, RCT	48	change in myocardial perfusion (SPECT)	negative	19
AGENT-3	CAD	Ad	FGF-4	percutaneous i.c. injections	phase III, RCT	416	ETT	negative (subgroup of >55-year-old women positive)	28
AGENT-4	CAD	Ad	FGF-4	percutaneous i.c. injections	phase III, RCT	116	ETT	negative	28
VEGF peripheral trial	PAD	Ad, Pl	VEGF-A165	percutaneous i.a. injection at angioplasty site	phase II, RCT	54	increased vascularity in angiography	positive	40
RAVE	PAD	Ad	VEGF-A121	i.m. injections	phase II, RCT	105	peak walking time	negative	42
WALK	PAD	Ad	HIF-1α/VP16	i.m. injections	phase III, RCT	289	peak walking time	negative	48
Delta-1	PAD	Pl	Del-1	i.m. injections	phase II, RCT	105	peak walking time	negative	47
Groningen trial	PAD	Pl	VEGF-A165	i.m. injections	phase II, RCT	54	decrease in amputation rate	negative	41
HGF-STAT	PAD	Pl	HGF	i.m. injections	phase II, RCT	104	limb perfusion measured as TcPO2	positive, TcPO2 increased in high-dose group at 6 months	53
HGF study	PAD	Pl	HGF	i.m. injections	phase II, RCT	40	improvement in rest pain, reduction in ulcer size	positive, primary endpoint improved 70%	54
TALISMAN	PAD	Pl	FGF-1	i.m. injections	phase II, RCT	125	ulcer healing	negative (secondary endpoint amputation rate positive)	44
TAMARIS	PAD	Pl	FGF-1	i.m. injections	phase III, RCT	525	time to major amputation or death	negative	45

CAD, coronary heart disease; PAD, peripheral vascular disease; RCT, randomized controlled trial; i.a., intra-arterial; i.c., intracoronary; i.m., intramuscular; i.my., intramyocardial; ECG, electrocardiogram; RV, retrovirus; Pl, plasmid; ETT, exercise tolerance test; hGCSF, human Granulocyte colony stimulating factor; TcPO₂, transcutaneous oxygen tension; ST, ECG S-T segment; NA, not available.

Adenoviral intracoronary delivery of FGF-4 has been thoroughly tested in a series of AGENT trials.²⁸ Currently, the ASPIRE trial will compare standard of care without a placebo group in an open-label design with a primary endpoint based on relative perfusion changes in the myocardium, as measured with ⁹⁹Tc-SPECT imaging at 8 weeks.²⁹ The randomized, double-blinded, placebo-controlled AWARE trial has also been planned with intracoronary adenovirus (Ad) FGF-4 in women with stable angina (Table 1). Both the ASPIRE and AWARE trials are based on significant positive effects in an exercise tolerance test found in AGENT 3 and 4 trials in women older than 55.²⁸ HGF has been used both as plasmid and adenoviral constructs in the treatment of coronary heart disease, but, so far, only small open-label studies have been reported.30, 31

Potential new treatments for refractory angina include upregulation of hypoxia-inducible factor 1α, thymosin B4, modified stabilized RNAs, GalNac-antisense oligonucleotides, nanoparticles, promoter-activating small interfering RNAs (siRNAs), and exosomes, which could deliver both growth factors and microRNAs (miRNAs) to ischemic myocardium.6, 32, 33, 34, 35, 36, 37, 38 No trials are currently testing these new approaches for the treatment of severe coronary heart disease. However, thymosin B4 is an interesting candidate because it activates partly different signaling cascades than previously used growth factors.³⁴ Also, epicardial activation of cardiomyocyte proliferation and tissue repair could prevent harmful remodeling and be a useful approach for therapy.³⁵

Peripheral Vascular Disease

Peripheral vascular disease is caused by atherosclerotic occlusions in the main lower extremity arteries, producing symptoms like claudication, rest pain, and insufficient ulcer healing. Patients are usually older than those with coronary symptoms, which makes this population very difficult to treat. Current pharmacological treatments are not very effective for critical limb ischemia or severe peripheral arterial disease, and not all patients are suitable for endovascular operations. Therapeutic vascular growth with the same factors used in coronary heart disease gene therapy has been tested with both intra-arterial and intramuscular delivery routes. However, in a recent meta-analysis of randomized controlled trials, no consistent benefits were found of angiogenic gene therapy in peripheral vascular disease.³⁹ This illustrates the challenges ahead for this population, but the driver for innovation is the large unmet clinical need. Some positive reports are available from both plasmid- and adenovirus-mediated VEGF-A gene therapy.⁴⁰ Also, improvements in oxygenation as well as walking time and ulcer healing have been reported,41, 42 and a plasmid-based VEGF-A product has been approved for clinical use in peripheral arterial disease in Russia.⁴³ Two large FGF-1 plasmid trials, TALISMAN⁴⁴ and TAMARIS,⁴⁵ did not find functional improvements in the treated patients, although the TALISMAN trial showed a beneficial effect on the amputation rate in critical ischemia patients. FGF-2 is currently tested in peripheral vascular disease (PAD) using a Sendai virus vector, but results have not been reported.⁴⁶ Of the other large previous trials, plasmid Del-1 (Delta-1 trial)⁴⁷ and adenoviral HIF-1α/vp16 (WALK trial),⁴⁸ aiming to stimulate angiogenic signaling pathways, gave negative results.

Problems like long atherosclerotic stenoses in the main arteries, shunting and a stealing effect of the increased blood flow to less ischemic areas, unstable neovessels, and too short a time of transgene expression have been listed as potential reasons for failure.⁴⁹ It appears that some basic physiological aspects of peripheral muscle perfusion have not been fully taken into account when designing previous trials, and it is likely that factors like overdilated capillaries, alterations in capillary blood transit time, and insufficient transfer of oxygen into peripheral muscles need to be taken into account when designing future clinical trials.6, 50 A recent trial aimed to improve the functionality of the induced neovessels by using a combination of retroviral VEGF-A and angiopoietin-1 gene transfers to endothelial and smooth muscle cells isolated from patients with peripheral vascular disease, whereafter the cells were injected back into the patients.⁵¹ A combination with surgical revascularization is tested in an ongoing KAT-PAD101 trial, where AdVEGF-D^dNdC is given 1–2 days prior to vascular surgery to dilate capillaries and small arterioles to improve distal runoff after the vascular surgery operation.⁵² In addition, AdVEGF-D^dNdC should induce both angiogenic and lymphangiogenic growth in the treated legs (Table 1). Two trials have tested HGF in critical limb ischemia,53, 54 and a current trial is ongoing using a plasmid that expresses two HGF isoforms. Multiple intramuscular injections at several time points are used for gene delivery. The SDF-1 plasmid is also tested in a similar setup for critical limb ischemia (Table 1).

It appears that, in the future, critical limb ischemia patients should be studied separately from less severe claudicants and that more emphasis should be given to accurate imaging and metabolic and functional studies as surrogates for the treatment effect because walking distance, amputation rate, or ulcer healing may not be able to capture improvements in the frequency of hospital admissions or related variables in these patient populations.⁵⁵ It is possible that only a subgroup of peripheral vascular disease patients will benefit from therapeutic vascular growth, whereas others may actually have problems other than insufficient angiogenesis.⁶ For future clinical trials, it would be important to identify patient subgroups that are most likely to respond positively to the new treatments.

Heart Failure and Arrhythmias

Heart failure is an increasingly common problem in the elderly population because of the improved therapies for coronary heart disease and acute myocardial infarction. Advances in the understanding of the pathogenesis of heart failure have led to recent gene therapy trials (Table 2).⁵ The CUPID2 trial tested adeno-associated virus (AAV)1-sarcoplasmic reticulum calcium ATPase 2a (SERCA2a) in patients with chronic systolic heart failure or non-ischemic cardiomyopathy.⁵⁶ Previous smaller trials had indicated positive effects; therefore, it was a great surprise that all endpoints in the CUPID2-trial were negative. The AAV1 vector was given with one-time intracoronary infusion, but, apparently, the gene transfer efficacy remained too low to induce any measurable positive effects.⁵⁷ The AGENT-HF and SERCA-LVAD trials will test the same AAV1-SERCA2a in congestive heart failure and in patients with a left ventricle assist device (LVAD), respectively (Table 2).

Table 2.

Ongoing, Planned, or Recent Trials in HF, Hyperlipidemias, Restenosis, or Vein Graft Disease

Trial	Disease	Vector	Therapeutic Agent	Delivery	Study Design	n	Primary Endpoint	Main Result	Reference
CUPID 2	heart failure	AAV1	SERCA2a	percutaneous i.c. injection	phase II, RCT	250	time to recurrent cardiovascular events	negative	56
SERCA-LVAD	chronic heart failure in patients with LVAD	AAV1	SERCA2a	percutaneous i.c. injection	phase II, RCT	24	safety and feasibility	NA	NCT00534703
AGENT-HF	heart failure	AAV1	SERCA2a	percutaneous i.c. injections	phase II, RCT	44	changes in left ventricular end-systolic volume	NA	NCT01966887
STOP-HF	heart failure	Pl	SDF-1	percutaneous i.my. helical infusion catheter-mediated injections	phase II, RCT	93	6-min. walking distance	negative	58
RETRO-HF	heart failure	Pl	SDF-1	percutaneous retrograde injection via coronary vein	phase I/II, open-label part and RCT part	52	6-min. walking distance	NA	NCT01961726
AC6	heart failure	Ad	adenylyl cyclase type 6	percutaneous i.c. injections	phase I/II, RCT	56	combined ETT and cardiac function before and during doputamine stress	NA	NCT00787059
Lp(a)	high lipoprotein Lp(a) levels	as-oligonucleotide	as-oligonucleotide against Lp(a)	weekly s.c. injections	phase II, RCT	64	change in plasma Lp(a) concentration	positive, 67% reduction in plasma Lp(a)	67
Lp(a)-L	high lipoprotein Lp(a) levels	GalNac-conjugated as-oligonucleotide	liver targeted as-oligonucleotide against Lp(a)	weekly s.c. injections	phase I/II, RCT	58	change in plasma Lp(a) concentration	positive, 66%–92% reduction in plasma Lp(a)	67
PCSK9	high LDL cholesterol levels	siRNA	siRNA against PCSK9 mRNA	weekly/monthly s.c. injections	phase I, RCT (healthy volunteers)	69	change in plasma LDL cholesterol	positive, plasma LDL cholesterol reduction 51%–60%	68
apoC-III	hypertriglyceridemia	as-oligonucleotide	as-oligonucleotide against apolipoprotein C-III	weekly s.c. injections	phase II, RCT	85	change in plasma apoC-III level	positive, apoC-III plasma levels decreased 40%-80%	69
AAV FH	homozygous familial hypercholesterolemia because of LDL receptor mutation	AAV	LDL receptor	liver-directed delivery	phase I, open-label	10	safety, decrease in plasma LDL cholesterol	NA	NCT02651675
Familial hypercholesterolemia trial	homozygous familial hypercholesterolemia because of LDL receptor mutation	RV	LDL receptor	ex vivo gene transfer to patient hepatocytes that were then injected back into the patient	phase I, open-label	5	safety, decrease in plasma LDL cholesterol level	moderate decrease in LDL cholesterol in three patients	2
Previous Trials in HF, Hyperlipidemias, Restenosis, or Vein Graft Disease
Glybera trials	lipoprotein lipase deficiency	AAV1	lipoprotein lipaseS447X	i.m. injections	phase I/II, open-label	19	post-Brandial hypertriglyceridemia, severe pancreatitis	positive, 50% decrease in pancreatitis, improvement in post-Brandial chylomicron metabolism	65, 66
Mipomersen trials	severe hypercholesterolemia	as-oligonucleotide	as-oligonucleotide against apolipoprotein B	weekly s.c. injections	phase III, RCT	158	change in plasma LDL cholesterol	positive, plasma LDL cholesterol reduction 37%	70
Italics	in-stent restenosis	as-oligonucleotide	as-oligonucleotide against c-myc	percutaneous i.c. local delivery after stent placement	phase II, RCT	85	percent neointimal volume obstruction by IVUS	negative	76
Prevent III	vein graft failure in PAD	oligonucleotide	E2F oligonucleotide decoy	ex vivo pressure-mediated delivery to vein graft	phase III, RCT	1,404	time to graft reintervention or amputation because of graft failure	negative	79
Prevent IV	vein graft failure in CABG	oligonucleotide	E2F oligonucleotide decoy	ex vivo pressure-mediated delivery to vein graft	phase III, RCT	2,400	angiographic vein graft failure	negative	80

as, anti-sense; FH, familial hyoercholesterolemia; IVUS, intravascular ultrasound; s.c., subcutaneous; CABG, coronary artery by-pass grafting.

The STOP-HF trial tested the SDF-1 plasmid in heart failure. Therapy was given as several endomyocardial injections using a catheter device. Six-minute walking distance was used as the primary endpoint, but this trial was also negative.⁵⁸ It appears that gene transfer efficiency was too low to achieve measurable positive outcomes. The RETRO-HF trial will test a similar SDF-1 plasmid using a retrograde delivery to the heart muscle. Adenoviral adenylyl cyclase 6 will be tested in congestive heart failure patients using a dose escalation format. The primary endpoints will include exercise tolerance tests and cardiac function measurements. However, the results of these trials are not yet available (Table 2). An obvious concern for the adenylyl cyclase trial is the short expression time of the adenovirus and potential immunological reactions, especially if repeated dosing will be required.

It is interesting to note that, even though heart failure is caused most commonly by advanced coronary heart disease, treatment genes for heart failure have been very different from those used for coronary heart disease and myocardial ischemia, targeting primarily excitation-contraction coupling and adverse remodeling. In the future, combination with short-term angiogenic therapies could help to balance altered cardiac metabolism and microcirculatory blood flow in failing myocardium. VEGF-B has shown several useful effects, such as cardiac-specific angiogenic activity and improvement in cardiac energy metabolism,59, 60, 61, 62 which could be tested in clinical trials. Combined gene and cell therapies could also potentially be used in future trials to achieve positive treatment outcomes.¹²

Heart failure develops gradually and, while in the early phase, targeting myocardial force, contractility, and adverse remodeling is useful for therapy; in the later stages, apoptosis of cardiomyocytes, fibrosis, and arrhythmias bring significant challenges for the late-state therapies. Because heart failure requires treatment of the majority of ventricular cardiomyocytes, very efficient vectors and gene delivery techniques are required for these indications. AAVs have shown significant long-term efficacy in preclinical models, but this still needs to be proven clinically. Also, the high prevalence of pre-existing antibodies against many AAV serotypes remains a concern. Recently, transfer of nucleic acids with slowly degrading nanoparticles and exosomes have become potential therapeutic approaches, and new treatment genes, such as S100A1, and vectors with better cardiac tropism are being developed for heart failure therapy.36, 37, 38, 63, 64 The best scenario would obviously be an efficient early treatment of myocardial infarction so that remodeling and progression to heart failure could be prevented. Improving blood flow to salvageable myocardium and evoking cardiomyocyte proliferation after acute myocardial infarction could be useful in this regard.

Arrhythmias are a significant cause of cardiac morbidity and mortality. Atrial fibrillation is a very common problem and can lead to significant clinical sequelae. The most severe ventricular arrhythmias and conduction system defects could theoretically be amenable to gene therapy. For example, a malfunctioning sino-atrial node could be replaced with a locally induced new focus, which could control cardiac rhythm. However, these approaches have turned out to be very difficult in vivo, and, so far, no clinical tests have been conducted in this area.

Hyperlipidemias, Lipoprotein Metabolism, and Atherosclerosis

Gene therapy could potentially be used for the treatment of severe inherited and acquired disorders of lipoprotein metabolism. The first approved gene drug in the Western world was Glybera, which is an AAV1 vector expressing lipoprotein lipase. This treatment is indicated for severe lipoprotein lipase deficiency, which causes high postprandial plasma triglyceride levels and severe pancreatitis attacks. The treatment is given by multiple intramuscular injections7, 65, 66 (Table 2).

Historically, homozygous familial hypercholesterolemia, which is caused by a mutation in low-density lipoprotein (LDL) receptor, was the first disorder where gene therapy was tested in the clinics.² The first trial was very laborious, with a large liver resection and harvesting of hepatocytes, followed by retrovirus-mediated ex vivo gene transfer of LDL receptor and return of the cells back to the patients. The results were not very convincing, although a few patients showed a moderate reduction in plasma cholesterol level. Currently, a clinical trial is planned with an AAV vector expressing LDL receptor to treat homozygous familiar hypercholesterolemia (Table 2). Based on good clinical experiences with AAV vectors in the treatment of hemophilia, this approach sounds promising, although potential immune responses against AAV vectors remain a concern.

Promising results have been obtained with stabilized, liver-targeted antisense oligonucleotides against the lipoprotein (Lp)a. Lp(a) is an independent risk factor for coronary heart disease, and subcutaneous injections of antisense oligonucleotides have been reported to be safe and very efficient in reducing atherogenic Lp(a) levels.⁶⁷ Other targets to reduce atherogenic lipoproteins are PCSK9⁶⁸ and apolipoprotein C-III,⁶⁹ which are currently tested for human therapy based on either antisense oligonucleotides or RNAi technology (Table 2). Mipomersen is already an approved antisense drug to reduce apolipoprotein B-100 levels in familial hypercholesterolemia and in some other types of severe hypercholesterolemias.⁷⁰ However, significant side effects have also been reported with mipomersen, and this drug is not used very often for therapy. Nevertheless, the liver seems to be an excellent target for antisense and RNAi therapy for lipid and lipoprotein disorders.

Some dyslipidemias and even atherosclerosis could be treated by increasing apolipoprotein A-1 levels⁷¹ because apoA-1 is a key component in antiatherogenic high-density lipoprotein (HDL) particles. Lecithin cholesterol acyl transferase is a key enzyme in HDL maturation and a potential target for therapy. Also, inhibition of microsomal triglyceride transfer proteins has shown potential to control severe atherogenic lipoprotein profiles. Because oxidized LDL is a key pathogenetic factor for the accumulation of lipids in the arterial wall,⁷² soluble decoy scavenger receptors, which can bind oxidized LDL and sequester it for degradation in the liver,⁷³ or vectors expressing antibodies against oxidized LDL⁷⁴ could theoretically be used as antiatherogenic therapy. However, it is clear that gene therapy approaches to control dyslipidemias and atherosclerosis can only potentially be used in very severe cases where well characterized and effective conventional therapies, such as statins and inhibitors of cholesterol absorption, cannot achieve the desired treatment effects.

Restenosis, In-stent Restenosis, and Vein Graft Disease

Intravascular interventions to target occlusions in the coronary arteries inevitably cause marked damage to the endothelium and arterial medial layer, exposing thrombogenic molecules. Activation of platelets, thrombosis, leukocyte adhesion, smooth muscle cell proliferation, matrix accumulation, and vascular remodeling are all involved in the pathogenesis of restenosis; i.e., the response to injury, leading to re-narrowing of the vessel lumen.⁷⁵ Drug-eluting stents have significantly improved treatment outcomes and reduced remodeling, recoil, and acute obstruction of the treated arteries, although late stent thrombosis still occurs in some patients. However, it is difficult to see how gene therapy could be competitive with current drug-eluting stents. Also, a previous clinical trial in restenosis gave negative results⁷⁶ (Table 2).

Vein graft stenosis is another common problem after bypass surgery. With most bypass operations still using the autologous saphenous vein as a conduit for bypass, there remains a significant risk of failure in the long term. The rationale for a gene therapy approach is to prevent pathological vein graft remodeling in the immediate time after implantation as the vein adapts to higher blood pressure and surgical handling. Vein grafts offer an excellent opportunity for local intraluminal or perivascular gene transfer during the preparation of grafts ex vivo.77, 78 Tissue inhibitor of metalloproteinases (TIMPs) have been suggested as a potential treatment,77, 78 but no clinical results are yet available from this approach because of problems in manufacturing clinical-grade AdTIMP-3. E2F transcription factor antisense decoys have been tried in clinics, but the results were disappointing79, 80 (Table 2). Vascular grafts and dialysis anastomosis stenosis could also be treated using gene transfer from the adventitial surface of the vessels. Phase I/II studies have indicated that local transduction with VEGF could reduce stenosis, presumably by increasing nitric oxide and prostacyclin production in the grafts. However, final clinical trial results are not available from these studies.⁸¹

Lessons Learned and Future Perspectives

Gene Delivery Vectors and Transduction Efficiency

It appears that the gene transfer efficiency in many previous cardiovascular trials has been too low to achieve meaningful clinical effects. Usually, in preclinical animal models, much higher doses per kilogram of gene transfer vectors have been used than what is possible in human trials.⁸² Intra-arterial delivery methods seem to be much less efficient than intramyocardial or intramuscular injections.⁴ As a consequence, it is likely that the concentration of therapeutic proteins in the target tissues has not reached sufficient levels and/or has not persisted long enough to achieve biological effects. Secreted therapeutic proteins have a better likelihood to give positive effects because, currently, it is very difficult to achieve more than 10%–20% transduction efficiency in human heart or peripheral muscles. Factors like volume per injection and matrix binding properties of the vectors and transgenes affect the distribution of transgene products in the treated tissues. It appears that physical (i.e., by surgery or via a catheter) or, possibly, genetic targeting of vectors would be desirable. In therapeutic vascular growth, the expression time of a transgene should be long enough to cause biological effects, but too long an expression time of powerful factors like VEGF will be detrimental to the tissue architecture.⁸³ In the future, it would be important to develop vectors where transgene expression could be regulated by small molecules or physiological stimuli. Integrating vectors can best be used for applications where genetic defects require a lifelong treatment effect. It would be important to determine, in preclinical studies and in phase 1 trials, that the transgene product can really be measured either in the plasma or in the target tissues and that a dose effect can be demonstrated.⁸⁴ Troubleshooting of the main problems in clinical gene therapy trials is presented in Table 3.

Table 3.

Troubleshooting of Clinical Cardiovascular Gene Therapy Trials

	Reasons	Potential Solutions
General Problems
Lack of clinical efficacy in randomized controlled trials	low gene transfer efficiency in target tissues (especially with naked plasmids and intra-arterial gene transfer routes)	intramyocardial or intramuscular gene transfer routes
	adenoviral or AAV dose too low in i.m. delivery	optimized viral dose, determine dose-response in the clinics
	short half-life of transduced therapeutic factors	enhance spreading of gene transfer vectors in the target tissues using sufficient injection volumes and number of injections
	difficult or unresponsive patient populations	use of gene transfer as adjuvant therapy in combination with conventional treatments
	strong placebo effects	use of less severely affected patients who can potentially respond to new therapies
		randomized, controlled, blinded study design
Specific Problems on Measured Variables
No detectable effects on clinical symptoms, organ function, metabolism, or imaging endpoints	gene transfer efficiency too low to cause physiological effects	perfusion measurements shortly after gene transfer to show proof of concept
	insufficient distribution of vector or transgene product because of binding to extracellular matrix	careful dose optimization in preclinical and early clinical studies
	regression of new vessels too fast after transient transgene expression	use of gene transfer vectors that produce long-term and/or regulated gene expression
	increased blood flow may cause shunting in target tissues	development and validation of more sensitive surrogate endpoints for imaging, organ function, and target tissue metabolism
	lack of validated, sensitive surrogate endpoints	understanding and controlling immunoresponses against vectors and transgenes
	wrong time points for endpoint measurements	combination of gene therapy with established treatment procedures
	immune responses against vectors or transgenes
	pathophysiology still incompletely understood

Patient Populations

In many previous trials, severe “no option” patients who are no longer eligible for any other kind of treatments have been recruited. However, in future ischemia and heart failure trials, healthier patients should be recruited. It is possible that, in severe ischemia patients, one of the main reasons for the difficult clinical situation is that endogenous angiogenesis has failed in the first place and that these patients are no longer able to react to angiogenic therapies. It is likely that only some subgroups of patients will respond positively to gene therapy approaches.⁸ It would be highly desirable to find biomarkers or other characteristics that would help to select the most optimal patients for future clinical trials.⁶

The placebo effect has been very strong in many previous cardiovascular gene therapy trials, and only a randomized, blinded, controlled trial design can give reliable results regarding clinical benefits.⁸² The reasons for the strong placebo effect are partly unknown, but patients usually go through a heavy pre-evaluation screening process with multiple laboratory and imaging tests, and gene therapy itself raises high expectations in severely affected patients. Therefore, it is not surprising that placebo groups have reacted positively. It should also be kept in mind that, in intramuscular injections, tissue injury caused by the needle or catheter can lead to local inflammation and production of growth factors and cause some positive effects. These factors should be taken into account when planning control groups for future clinical trials (Table 3).

Endpoints

Whether technical or pharmacological shortcomings of gene drugs in previous trials have caused clinical failures remains unknown. However, traditional endpoints, such as survival, exercise tolerance, amputation frequency, or ulcer healing, seem to be very demanding in severely affected patients.⁶ There is a clear need to develop validated surrogate endpoints for cardiovascular trials based on tissue perfusion, collateral flow, metabolic improvements, and reduced burden of the use of hospital services that could better capture potential therapeutic benefits. This, together with selection of patients who are most likely to benefit from new therapies, should be taken into account in future clinical trials.6, 8

An evaluation of pre-existing antibody levels and immunological responses to the treatments should be included in every trial and correlated with the potential treatment results. In future trials, gene therapies should also be combined with existing therapies, such as bypass surgery or angioplasty as adjuvant therapies. This would allow new approaches for therapy, such as in peripheral vascular disease, where gene transfer is given prior to bypass operation to open peripheral capillaries and to avoid a poor run-off syndrome, which often leads to disappointing results in these patients.6, 50, 52 Other co-morbidities can significantly affect therapeutic outcomes, and, for this reason, patients should be accurately stratified for diseases like diabetes, rheumatoid arthritis, or other autoimmune diseases to avoid confounding results. Most therapeutic approaches have so far been based on single-dose applications. It should be evaluated whether repeated dosing could be applied because many chronic cardiovascular diseases are likely to progress, and the need for future therapeutic interventions is quite obvious.

As far as the safety profile of cardiovascular gene therapy is concerned, most trials have shown a very good safety profile even after 10-year follow-up.23, 24, 25 This should encourage clinical testing of new therapeutic approaches with improved vectors and gene delivery methods. Current imaging methods can be used to guide therapeutic approaches so that very low doses could be applied to targets like ischemic myocardium in combination with standard therapies. Because a regulatory pathway for the approval of gene drugs has now been established, it is expected that several novel treatments for cardiovascular diseases will enter clinical testing in the near future.