Vipul R. Panchal, MD is a Fellow in Interventional Cardiology at the Krannert Institute of Cardiology, Indiana University School of Medicine, Indianapolis, IN. He completed his General Cardiology Fellowship and his Internal Medicine Residency at the same institution. He received his MD from Northeastern Ohio Universities College of Medicine in 1996.

Vijay Kalaria, MD is an Assistant Professor of Clinical Medicine and an Associate Director of Cardiac Catheterization Laboratories and Interventions, Krannert Institute of Cardiology, Indiana University School of Medicine, Indianapolis, IN.

Jeffrey A. Breall, MD, PhD is a Clinical Professor of Medicine and the Director of Cardiac Catheterization Laboratories and Interventions, Krannert Institute of Cardiology, Department of Medicine, Indiana University School of Medicine, Indianapolis, IN.

Keith L. March, MD, PhD is an Associate Professor of Medicine at the Krannert Institute of Cardiology, Department of Medicine, Indiana University School of Medicine, Indianapolis, IN.

Myocardial gene therapy is an emerging therapeutic option for the treatment of patients with advanced ischemic heart disease. Early clinical experiences show feasibility and potential efficacy of gene-mediated therapeutic angiogenesis. As percutaneous methods of gene delivery mature, the interventional cardiologist will undoubtedly play a major role in the delivery of these agents and will require a working knowledge of this field. Future placebo-controlled, large, randomized trials with both current and new angiogenic factors will define the ultimate role of this therapy.

Patients with coronary artery disease refractory to medical therapy and without revascularization options continue to challenge the cardiologist. In recent years, myocardial gene therapy for the treatment of heart disease has evolved from bench to bedside. Early clinical trials of myocardial angiogenesis have shown promise ¹ and emphasize the need for further studies to determine the optimal delivery method, agent, and dose for safe and effective therapy.

This review will focus on current aspects of catheter-based gene therapy for myocardial angiogenesis that are relevant to the interventional cardiologist, including possible delivery methods, agents, vectors, and potential problems with each strategy.

Background

The prevalence of symptomatic patients with unrevascularizable end-stage coronary artery disease is increasing. Many of these patients remain symptomatic despite maximal anti-anginal therapy and may not be amenable to or may have failed surgical or percutaneous forms of myocardial revascularization. ² It is estimated that approximately 12% of symptomatic patients with coronary disease undergoing cardiac catheterization are not candidates for percutaneous coronary intervention or coronary artery bypass grafting (CABG). These patients suffer high rates of cardiac events and have limited therapeutic options. ³ A potential therapeutic option for these patients is therapeutic angiogenesis.

Therapeutic angiogenesis

Therapeutic angiogenesis is an intervention aimed at inducing neovascularization and promoting collateral formation in ischemic myocardium through the administration of angiogenic growth factors or their genes. ⁴ It is well known that coronary collateralization improves the prognosis of patients with obstructive coronary artery disease by limiting infarction size, preserving ventricular function, and enhancing overall perfusion in ischemic myocardium. ^5,6 Thus, augmentation of collateral vessel growth may be a valuable approach in these patients.

Angiogenesis is a process in which new capillaries sprout from the existing vasculature by way of proliferation and migration of endothelial cells. This is mediated, in part, by angiogenic growth factors. ^1,7 Wound healing and reproductive growth are examples of physiologically normal forms of angiogenesis. Pathologic forms include tumors and diabetic retinopathy.

Angiogenesis can also be a response to ischemia. It is suggested that patients with extensive coronary disease may have an impaired endogenous angiogenic response, and enhancement of this process is the basis of therapeutic angiogenesis. ¹

The feasibility of therapeutic myocardial angiogenesis was shown in initial animal studies, primarily using the angiogenic proteins, such as vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF). Early attempts to enhance collateral perfusion in ischemic myocardium used local administration of recombinant forms of these proteins. ⁸ These studies showed the efficacy of exogenous supplementation of growth factors by demonstrating enhancement of collateral flow as well as improvement in myocardial perfusion and in wall motion. ^9-11

Rationale for gene therapy

Administration of recombinant proteins for angiogenesis has shown benefits in several phase I trials. ¹² However, there are several limitations to protein delivery methods. A critical problem is inadequate myocardial exposure time due to brief serum and tissue half-life of the protein. Higher initial concentrations and multiple administrations of the angiogen would provide efficacy, but at the expense of potential side effects. ⁷ In contrast, gene therapy modalities offer a sustained, but self-limited, expression of therapeutic factors that can achieve desired biological effects in a single dose. ¹

Gene transfer

The goal of cardiac gene therapy is to deliver genetic material (frequently by plasmid or viral vectors) to the myocardium in order to induce sustained and local overexpression of the transgene. This process is called transfection. ¹³ The degree of physiologic effect derived from gene transfer is dependent on the efficiency of the transfection process. Because of variable transfection efficiencies, gene therapy has a more unpredictable dose-response relationship in comparison with protein methods. ⁷

Successful gene therapy is contingent on several consecutive steps (Figure 1): (1) the method of gene delivery to the target tissue, (2) a capable vector system allowing intracellular passage of genetic material (plasmid or virus), and (3) robust local expression of an appropriate transfected gene (angiogenic agent) within the target tissue. ¹³

Delivery method

Various routes of administration have been used in delivering angiogenic proteins and genes. ⁸ The optimal technique maximizing safety and efficacy for myocardial gene delivery is not well established.

An ideal technique would deliver angiogenic factors in a site-specific manner to the target tissue, minimizing systemic exposure. It should achieve high local concentrations as well as provide sufficient tissue retention time to allow cell interaction and subsequent gene transfer. ¹² Finally, the technique should be simple to perform, relatively inexpensive, and use existing technologies.

Catheter-based myocardial delivery can be divided into three methods: intravascular, intramyocardial, and intrapericardial.

Intravascular approaches

Intravenous (IV)-- Intravenous administration is the simplest form of delivery. Its major drawback, however, is very low myocardial tissue uptake. In radiolabeled distribution studies, only 0.26% of bFGF was isolated in ischemic myocardium 1 hour after IV injection and 0.04% after 24 hours. ¹⁴ This results in significant systemic exposure of the agent, consequently increasing the risk for side effects. Accordingly, IV administration of agents will not be effective in inducing angiogenesis. ⁷

Intravenous administration of an agent that is designed with a "homing" mechanism can improve myocardial delivery. This has been performed by complexing therapeutic agents with ultrasound contrast bubbles that can be injected intravenously and subsequently lysed at the myocardium with transthoracic high-frequency ultrasound. ¹⁵

Intracoronary -- The intracoronary route has been used in many clinical trials of angiogenesis. ^1,16-18 It is favored because it is relatively simple to administer in the catheterization laboratory, requires no special equipment, is widely available, and is suited to most patients with coronary disease. ⁷ The intracoronary route is limited by poor myocardial tissue uptake, although it is slightly better than the IV route. Approximately 3% to 5% of bFGF was recovered from the myocardium several hours after intracoronary injection, and with a low 0.04% retention after 24 hours. ¹⁶ In a phase I study, patients receiving selective intracoronary injections of FGF-2 incurred similar systemic exposures to those receiving IV infusions. ¹⁶ Despite relatively poor first-pass uptake, animal studies and phase I studies seemed to indicate potential efficacy of single bolus intracoronary protein injections. ^1,10,11

Coronary venous retroinfusion-- Coronary venous retroinfusion is gaining interest as an efficient myocardial gene-delivery modality. Retroinfusion into the coronary vein is very effective in delivering drugs to the myocardium and microvascular bed. ¹⁹ Deposition is enhanced significantly in ischemic myocardium during balloon occlusion of a corresponding coronary artery. Significantly longer tissue retention has also been reported by this route. ²⁰

This method is performed by cannulation of a selected cardiac vein via the coronary sinus using a modified balloon-tipped perfusion catheter. The coronary vein is then occluded to prevent drainage. Pressure-regulated infusion of the gene agent is performed at an internal venous pressure of 20 mm Hg for several minutes. ²⁰ Another
modification of this technique is selective high pressure (100 mm Hg)
single-bolus injection (10 mL). The technique appears to be safe because no echocardiographic evidence of myocardial dysfunction has been noted, and myocardial marker release is minimal. ²¹

Gene delivery with this method has been shown to be feasible. Pressurized retrograde coronary venous delivery of an adenovirus has shown considerably more gene expression than that of antegrade intracoronary administration in pigs. ²⁰ The technique is also effective for plasmid-driven angiogenesis as well. Administration of plasmid developmental endothelial locus-1 (del-1, an angiogenic factor) improved myocardial perfusion in ischemic pigs. ²² In addition to gene delivery, coronary venous retroinfusion provides widespread distribution of cells and microspheres to the myocardial interstitium. ²³

Intramyocardial approaches

Until recently, all myocardial gene delivery methods were surgical in nature. Early phase I clinical trials of gene-based myocardial angiogenesis administered growth factors by direct transepicardial injection at the time of CABG or using minithoracotomy. ^1,7 This approach was favored over intravascular methods due to the lack of significant myocardial washout and dissemination to distant tissues after initial injection. ⁸ Subsequent animal studies have shown, however, that incomplete retention of injectate also occurs with intramyocardial delivery, although to a lesser degree. ²⁴

Transendocardial intramyocardial-- Currently, the only clinically feasible catheter-based method of intramyocardial administration is transendocardial intramyocardial delivery via an intracavitary left ventricular approach. This technique has been validated in preclinical and clinical studies. ²⁵ In a biodistribution study comparing epicardial open chest injection of microspheres with percutaneous endomyocardial delivery in pigs, the endomyocardial tissue retention was better (43% versus 15%) than thoracotomy-based epicardial injection. ²⁴

Transendocardial injections are performed with various specialized catheters that have retractable sheathed needles ²⁵ or helical infusion needles. ²⁶ The catheter is advanced intra-arterially across the aortic valve into the left ventricle and endocardial injections are performed. There are several adaptations to this technique, using different guidance methods such as intracardiac echocardiography (ICE), ²⁷ fluoroscopy, ²⁸ and electromechanical mapping. ^25,29 The only published clinical experience to date is with electromechanical guidance. ^25,29

Fluoroscopic-guided endocardial gene transfer alone is not likely to be useful clinically because it does not provide adequate anatomic information about injection sites or the nature of the myocardium. However, ICE guidance in conjunction with fluoroscopy provides useful information, such as directing needle position and monitoring systemic leakage of gene injectate, and it provides the ability to see injected sites to avoid repeat administration. ²⁷

Transendocardial injection guided by a left ventricular electromechanical mapping system (NOGA, Biosense Webster, Inc.; Diamond Bar, CA) has gained considerable interest. Using a magnetic field, the system is able to precisely locate the catheter in the ventricle in three dimensions. In addition, electromechanical mapping can delineate normal, ischemic, or hibernating myocardium based on the amplitude of unipolar endocardial potentials and assessment of local myocardial cell shortenening. ^25,29

Disadvantages of this technique include the requirement of multiple injection sites, relatively small injection volumes (0.1 to 1 mL), and a relatively complex mapping method.

Intrapericardial approach

Another catheter-based method that provides efficient gene transfer is intrapericardial administration. Catheters have been developed to access the pericardial space percutaneously in a safe manner in animals. ³⁰ A helical-tipped transventricular penetrating catheter has been used to instill various genes into the pericardial sac of dogs. Under fluoroscopic guidance, the catheter is advanced into the right ventricle, and the tip is passed through the apex. Alternatively, a percutaneous transthoracic catheter incorporates a sheathed needle with a suction device to grasp the pericardium and access the pericardial space.

Robust gene expression has been seen in the pericardium and epicardium with these techniques, while serum and distant tissue levels were undetectable or at background levels. ³¹ Minimal systemic exposure, high transgene expression from adenoviral vectors, and potentially widespread myocardial delivery are valuable characteristics. However, the clinical use of this technique may be limited because of prior pericardial manipulation in many patients with cardiac surgery. ⁸ Furthermore, assessment of theoretical risks such as intrapericardial bleeding with angiogenesis therapy needs further study. ¹³

Gene transfer vectors

Cardiovascular gene transfer requires effective and durable vectors that carry and facilitate entry of genetic material into the target cell and to the nucleus. ¹³ Once this is accomplished, sustained expression of the transgene can occur. The vector usually contains a promoter region that allows for expression and translation of the gene, a copy DNA segment encoding the angiogenic factor, and various signals for ending translation and for posttranslational processing. ³² The traditional types of vectors used for gene therapy are nonviral and viral.

Nonviral gene transfer

Nonviral plasmid-mediated DNA transfer is used because of its safety profile and transient expression. Many viral vectors produce an immune response to some degree, which is avoided with nonviral methods. In addition, transient transgene expression occurs in the nucleus without incorporation into chromosomal DNA. ¹³ Plasmids can also incorporate larger cassettes to code multiple genes or larger proteins. A major disadvantage, however, is that naked DNA has a short half-life due to degradation by circulating nucleases and scavenger receptors. ³³ Plasmids are also very inefficient in transfection. ³⁴

The formulation of plasmid DNA complexes with macromolecules has enhanced stability and transfection. ³⁴ The DNA is combined with polymers, such as cationic liposomes or poloxamers. Negatively charged DNA combines with positively charged cationic lipids that form stable structures, which isolate plasmids from nucleases and facilitate entry into the cell by endocytosis. ³⁴

Another novel way to enhance and target nonviral DNA delivery is with the use of microbubbles. Plasmid DNA is complexed with microbubbles, delivered to the myocardial circulation, and subsequently disrupted with high-power ultrasound releasing the genetic material to the target area. It is based on contrast microbubbles similar to those used in contrast echocardiography and can be formulated of albumin, polymeric shells, or cationic lipid membranes. ¹⁵ In a porcine study of systemically delivered plasmid DNA, transfection occurred only in myocardium that was exposed to the ultrasound beam. ³⁵ Similarly, a 10-fold increase in transfection efficiency was noted with an adenovirus formulated with microbubbles. There was, however, extensive transgene expression in the livers of all animals in which this method was used, ¹⁷ suggesting that site-specific delivery may still be necessary.

Viral gene transfer

Gene transfer with the use of engineered viruses is very efficient in transfecting cells. This is primarily due to the innate ability of a virus to infect a cell and transfer its genetic material. Viruses are not subject to intravascular degradation in contrast to plasmids. Several different types of viruses can be used for gene therapy. The concerns of viral gene therapy are of host immune responses and prolonged gene expression leading to overexposure to the angiogenic factor. ³⁶

Adenoviruses have been studied extensively and provide short-term, 1- to 2-week, transgene expression without integrating into host DNA. They are capable of transfecting both the proliferating and nonproliferating cells. These viruses are rendered replication-deficient, and the genome is replaced with the desired angiogenic expression cassette. ¹³ Newer generation adenoviruses have a longer duration of expression and less immunogenicity. Adeno-associated viruses provide longer duration of expression as well. ¹ Retroviral vectors transfer RNA, which is transcribed to DNA and integrated into the host genome resulting in long-term expression. Retroviruses have limited use in the myocardium, as they can only transfect proliferating cells. ³²

Angiogenic agents

Many angiogenic growth factors have been identified (Table 1). ^12,37 However, all clinical trials of angiogenesis to date have studied the VEGF and FGF classes. ¹² Human cardiovascular gene therapy trials in progress or in early development are now expanding to trials using platelet-derived growth factor, hypoxia inducible growth factor-1*, del-1, and nitric oxide synthase. ³⁷

Vascular endothelial growth factor is an endothelial cell-specific mitogen and exerts its influence by means of specific receptors. ¹ There are several isoforms of the VEGF protein and all isoforms impart angiogenic effects. ³² However, of the several receptors for VEGF, only the VEGFR-2 (KDR) receptor mediates angiogenesis.

Fibroblast growth factors are a family of several angiogens that share a similar protein structure. In contrast to VEGF, their effects include proliferation of multiple cell types including endothelial cells, smooth muscle cells, and fibroblasts. ³⁸ Both VEGF and FGF can enhance production of various proteases essential to matrix degradation for angiogenesis. ^32,38

Safety concerns of gene therapy

Despite the promise seen in both preclinical and early clinical gene transfer studies, concerns regarding safety of gene delivery and toxicity of therapy have arisen. The death of a patient in a phase I gene therapy trial for treatment of ornithine transcarbamylase deficiency due to an adenovirus-mediated reaction has triggered concern. ³⁶ Adenoviral vectors are known to induce inflammatory and immunologic reactions that are sometimes severe. Furthermore, VEGF trials have reported hypotension, edema, and formation of spider angiomas. ^1,39 In some cases, overexpression of VEGF has produced vascular tumors, vessel clusters, and severe localized edema. ¹ Similarly, FGF has shown potential toxicities including hypotension, proteinuria, and thrombocytopenia. ^13,16

The risk of pathologic angiogenesis is also of concern due to the possibility of enhanced growth of clinically silent malignancies. Retinal neovascularization (diabetic retinopathy and retinal vasculopathies) is associated with elevated levels of VEGF. ¹²

Trials of catheter-based angiogenesis

Initial catheter-based phase I clinical studies administered recombinant protein by the intracoronary route. These studies were small and without placebo controls, but did show potential benefit including improvements in exercise duration, perfusion, and fewer anginal symptoms. ¹ Larger phase II placebo-controlled clinical studies with the same approach, however, did not yield the same results. Patients treated with intracoronary and IV VEGF in the VEGF in Ischemia for Vascular Angiogenesis Trial failed to meet endpoints of improved exercise time and angina class. ¹⁸ In the FGF-2 Initiating Revascularization Support Trial, intracoronary injection of FGF-2 did not improve exercise time or improve perfusion. ¹ These findings are not surprising given the unfavorable pharmacokinetics of intracoronary delivery described previously.

To date, there are three published clinical trials of catheter-based gene transfer for angiogenesis (Table 2). ^17,25,29 These trials evolved from the need for a placebo control arm. Early phase I gene trials used intraoperative direct myocardial injection for gene transfer that precluded randomization to placebo due to the inherent risks of surgery and anesthesia. ¹⁷ Percutaneous methods allowed a control arm to the study design.

The first percutaneous trial of catheter-based gene therapy was a pilot study to show the safety and feasibility of transendocardial gene transfer using left ventricular electromechanical mapping. Vale and colleagues ²⁵ randomized 6 patients to plasmid VEGF-2 gene transfer or a placebo procedure. Improvements in anginal episodes and nitroglycerin use were noted in both groups, though the placebo group later experienced worsening angina, whereas the patients in the gene transfer group continued to improve. Perfusion scores and electromechanical function in ischemic sites were significantly better in the VEGF-treatment group than in the placebo group.

Based on these initial findings, a phase II dose-escalation study of 19 patients using the same method was performed and significant improvements were found in the Canadian Cardiovascular Society anginal class (3.5 versus 2.2, P = 0.012) and a significant mean change in angina class for patients treated with plasmid VEGF-2 (­1.2 versus 0.1, P = 0.04). There were trends toward improvements in exercise duration, angina questionnaire data, and perfusion on single-photon emission tomography. ²⁹

Finally, in the Angiogenic GENe Therapy Trial, a phase II, double-blinded, randomized, placebo-controlled trial of 79 patients showed safety of intracoronary adenoviral gene transfer of FGF-4. Treated patients showed a nonsignificant increase in exercise duration at 4 weeks (1.3 versus 0.7 min, P = not significant). However, in a subgroup analysis of patients excluded with exercise times >10 minutes, this difference became significant (1.6 versus 0.6 min, P = 0.01). ¹⁷

It is important to note the limitations of these and other trials of gene therapy. The studies are relatively small, and many are nonrandomized or uncontrolled. Furthermore, endpoints such as exercise duration and health status measures are "soft" endpoints and can be subjective. ⁷ These issues underscore the need for larger phase II, randomized, placebo-controlled trials to definitively prove efficacy.

Conclusion

The application of percutaneous catheter-based gene therapy techniques will allow larger scale studies to be performed. These same techniques may apply to other emerging technologies (such as cell delivery/transplantation for myocardial regeneration), thus continuing to expand the role of interventional cardiologists.