authors:
1Hong H. Keo, MD, 2Alan T. Hirsch, MD, 1Iris Baumgartner, MD, 3Sigrid Nikol, MD,
2Timothy D. Henry, MD

________________________________

Abstract

Critical limb ischemia (CLI) represents the most severe stage of atherosclerotic lower extremity peripheral artery disease (PAD), and CLI prevalence is expected to increase as the population ages. The current standard of care for CLI relies on direct revascularization, either by endovascular techniques or open surgical approaches, as there are few effective medical treatments for this condition. Therapeutic angiogenesis is a novel approach to improving limb outcomes for these patients. Experimental preclinical studies and phase I/II clinical trials of therapeutic angiogenesis using gene transfer in patients with CLI unsuitable for revascularization have shown promising results. In this review, we describe the potential clinical impact of this new approach as an adjunct in our therapeutic armamentarium.

Introduction

Critical limb ischemia (CLI) represents the most advanced stage of atherosclerotic, lower extremity peripheral artery disease (PAD) and is associated with high rates of cardiovascular morbidity, mortality, and major amputation.1 The incidence of CLI is estimated to be 125,000 to 250,000 patients per year in the United States and is expected to grow as the population ages.2–5 The 1-year mortality rate of patients with CLI is 25% and may be as high as 45% in those who have undergone amputation.1,6 The current standard of care for individuals with CLI includes lower extremity revascularization, either through open peripheral surgical procedures, endovascular techniques, or lower extremity amputation (i.e., if revascularization has failed or is unfeasible).1,7

Despite advanced techniques in endovascular and surgical procedures, a considerable proportion of patients with CLI are not suitable for revascularization. Of these patients, 30% will require major amputation and 23% will die within 3 months.8

Therapeutic angiogenesis is a novel strategy under investigation for the treatment of PAD that utilizes angiogenic growth factors, genes to encode these growth factors, or stem cells to promote neovascularization, in an attempt to increase perfusion to ischemic tissues through various mechanisms of action.9 Early clinical trials of gene transfer for therapeutic angiogenesis have been promising and provide hope for CLI patients who are unsuitable candidates for revascularization.

This paper will review the impact of therapeutic angiogenesis for the treatment of patients with CLI, including the safety and efficacy data provided by those clinical trials completed to date, and will outline potential future directions for clinical research.

The Concepts Underlying Therapeutic Angiogenesis

The concept of therapeutic angiogenesis evolved from pioneering work by Folkman,10 who observed that the development and maintenance of an adequate microvascular supply is essential for the growth of neoplastic tissue. Following the early identification of angiogenic growth factors, cardiovascular investigators began testing the hypothesis that stimulating angiogenesis could improve perfusion and function in ischemic tissues independent of macrovessel manipulation.11 Therapeutic angiogenesis involves the administration of angiogenic growth factors, as recombinant protein or gene encoding for those growth factors or stem cells to augment the collateral circulation and enhance blood flow to ischemic tissues. Angiogenic protein growth factors have been utilized, but require intra-arterial delivery and have short half lives.12,13 In particular, for PAD, gene therapy has theoretical advantages of intramuscular delivery, permitting repeated administration and a more prolonged therapeutic effect.

Angiogenic growth factors can be administered using non-viral or viral-vector encoding genes. The non-viral method uses naked plasmid DNA to transfer the gene encoding the desired angiogenic protein to the ischemic tissue.

The viral delivery method uses viruses as a vector to introduce the new gene to the ischemic tissues, also referred to as transfection. The viruses most often used in angiogenic studies are adenoviruses.14 The major advantage of this method is that transfection of growth factors can be achieved with high efficiency. However, the potential downside of this method is that transfection efficiency may be limited by prior viral exposure, and repeated administration might not be efficacious, due to rapid degradation of the viral vector.15,16

Therapeutic angiogenesis and angiogenic growth factors. Therapeutic angiogenesis was first evaluated by Dr. Jeffrey Isner in a 71-year-old patient with severe PAD and great toe gangrene in 1994.17 Human plasmid phVEGF165 in a dose of 2 mg was applied to the hydrogel polymer coating of an angioplasty balloon. By inflating the balloon in the vascular lumen, plasmid DNA was transferred to the distal popliteal artery.17 Functional and angiographic parameters improved within 12 weeks, and spider angiomata and edema developed unilaterally in the affected limb, suggesting the treatment had a local angiogenic effect. Since this pioneering, experimental therapy numerous angiogenic growth factors have been developed and tested in clinical trials with demonstration of angiogenic potential.18 Table 1 provides an overview of angiogenic growth factors that have been identified to stimulate neovascularization in preclinical and clinical models. Of these, the following four have been thus far evaluated in clinical trials of patients with CLI.

The most potent angiogenic factor affecting endothelial cell proliferation is vascular endothelial growth factor (VEGF), also known as vascular permeability factor. This factor has a high affinity for binding to endothelial cells. It was first discovered in the early 1980s as an agent that triggers hyperpermeability and macromolecule extravasation from microvessels.19 Four different receptors are known to bind the members of the VEGF family: VEGFR-1, VEGFR-2, VEGFR-3, and neuropilin-1.20 Endothelial cell specificity has been considered to represent an important advantage of VEGF for therapeutic angiogenesis because endothelial cells represent the critical cellular element responsible for new collateral vessel formation.21

Fibroblast growth factor (FGF) consists of 23 structurally related members and is important for the growth and migration of many cell types in the vessel wall, including endothelial cells and smooth muscle cells.22, 23 FGF is a key regulator of vessel growth and has successfully been used to promote angiogenesis.24

Hypoxia-inducible factor 1 (HIF-1) functions as a master regulator of oxygen and undergoes conformation changes in response to oxygen concentration. It is a heterodimeric transcription factor complex that consists of two subunits, HIF-1a and HIF-1.25,26 HIF-1aβ is expressed in the cell nucleus and its activity is not controlled by oxygen homeostasis, whereas the activity of HIF-1 is regulated by oxygen levels.27 Human endothelial cells transfected with Ad2/HIF-1a/VP16 (Ad, adenovirus; VP16, herpes simplex virus VP transaction domain) have been shown to promote endothelial proliferation and tube formation as a result of up-regulation of the expression of multiple angiogenic factors.28

Hepatocyte growth factor (HGF) is a mesenchymal-derived pleiotropic factor that regulates cell growth, cell motility, and morphogenesis of various types of cells. HGF is also a powerful angiogenic growth factor and is secreted by vascular endothelial cells and smooth muscle cells.29

Angiogenic Activity and Clinical Efficacy

Vascular endothelial growth factors. After introducing feasibility of angiogenic gene transfer of VEGF165 using human plasmid as a vector in 1994,17 Isner and coworkers30 conducted a non-randomized phase I clinical trial to document the safety of intramuscular phVEGF165 gene transfer in 6 patients (7 limbs) with CLI due to Buerger’s disease and who were not candidates for catheter or surgical revascularization. VEGF165 plasmid DNA was administered into the calf or distal thigh muscles with doses of 2 x 2 mg. After 14-month follow up, evidence of improved perfusion to the distal ischemic limb was documented by an increase in ankle-brachial index (ABI) in 3 limbs and by magnetic resonance imaging (MRI) in all 7 limbs, as well as by newly visible collateral vessels formation shown on serial contrast angiography in 7 of the 7 limbs.30 In another phase I study, 9 patients (10 limbs) received intramuscular injection of phVEGF165 into the ischemic muscle in a dose of 2 mg and repeated similar injection was performed after 4 weeks of initial injection.21 After 6-month follow up, evidence of improved perfusion was documented by increased ABI values, contrast angiography revealing newly visible collateral blood vessel formation, improved distal flow on magnetic resonance angiography, and ulcer healing and limb salvage in 3 patients who would have been amputated otherwise.21

Shyu et al31 conducted a phase I clinical study of 21 patients (24 limbs) with CLI unsuitable for catheter or surgical revascularization using plasmid DNA to deliver VEGF165 into the ischemic muscle in a dose escalation of 0.4, 0.8, 1.2, 1.6, and 2.0 mg at baseline and at 4 weeks thereafter. After 6-month follow up, evidence of improved perfusion was documented by ABI values (0.58 ± 0.24 to 0.72 ± 0.28; p < 0.001), and MRI revealing improved distal flow in 19 limbs (79%). Ischemic ulcers healed or improved markedly in 12 limbs (75%), and rest pain was relieved or reduced markedly in 20 limbs (83%). Amputation was performed in two patients because of wound infection.31

In a double-blind, randomized, placebo-controlled trial, 54 adult diabetic patients with CLI and who were unsuitable for revascularization were assigned to receive either intramuscular injection of placebo (n = 27) or VEGF165 (n = 27) in a dose of 2 mg at baseline and 4 weeks thereafter using plasmid as a vector.32 After 3-month follow up, the primary endpoint of amputation was decreased (3 in VEGF165 vs. 6 in placebo), but not statistically significant. However, improvement in skin ulcers (7 vs. 0; p = 0.01), decrease in rest pain (5 vs. 2), and an overall beneficial response (14 vs. 3; p = 0.003) was documented as compared to placebo.32

In addition to use of naked plasmid DNA constructs, the VEGF165 gene has been delivered using adenoviral vectors. In a phase II double-blind, placebo-controlled trial, 54 patients with intermittent claudication (n = 40) or CLI (n = 14) were assigned to receive either placebo, VEGF165 using an adenoviral vector (VEGF-Ad) or VEGF165 in a plasmid/liposome transfectant (VEGF-P/L) to deliver the angiogenic gene into the ischemic muscle.33 After 3-month follow up, vascularity documented by digital subtraction angiography was significantly increased in the VEGF-Ad and VEGF-P/L groups compared to placebo.33

Fibroblast growth factor. In a phase I study, 51 patients with unreconstructable severe PAD with rest pain or tissue necrosis underwent treatment with intramuscular injection of non-viral FGF (NV1FGF).34,35 Increasing single (0.5, 1, 2, 4, 8, and 16 mg) and repeated (2 x 0.5, 2 x 1, 2 x 2, 2 x 4, and 2 x 8 mg) doses of NV1FGF were injected into the ischemic thigh and calf using plasmid as a vector. NV1FGF was well tolerated, and after 6-month follow up of 38 patients, a significant reduction in pain scale and ischemic ulcer size, as well as increase in TcPO2 compared to baseline values was observed.35

In a randomized trial of 6 patients with CLI who were scheduled to undergo major amputation, patients were assigned to receive NV1FGF 3 to 5 days prior to amputation.16 NV1FGF was administered at 8 intramuscular sites with doses of 0.4 to 4 mg. This trial documented FGF1 transgene expression at all doses up to 3 cm from the injection sites, and that disseminated plasmid into blood vessel was rapidly degraded.16 In the phase II double-blind, placebo-controlled, multicenter trial conducted in the USA, 71 patients with CLI were assigned to receive placebo or 1 of 5 treatment regimens of 2 to 16 mg of NV1FGF delivered via 8 intramuscular injections in the affected leg.36 This trial showed that NV1FGF up to 16 mg was safe and well tolerated and that the primary endpoint of TcPO2 was increased in both NV1FGF- and placebo-treated patients, but there was improvement in ulcer healing in the NV1FGF-treated patients.36

In the recently published TALISMAN trial9, a phase II double-blind, placebo-controlled, multicenter study conducted in Europe, 125 patients with CLI unsuitable for catheter-based or surgical revascularization were assigned to receive intramuscular injection of either placebo or non-viral FGF (NV1FGF). Eight intramuscular injection sites were used to administer 4 mg of NV1FGF and similar injections were repeated after 15, 30, and 45 days. After 6-month follow up, the primary endpoint of complete ischemic ulcer healing was similar in both groups (19.6% vs. 14.3%), but all amputation, major amputation, and the combination of major amputation and death were significantly reduced in the intervention group (Figure 1).9

In another phase I/II, double-blind, placebo-controlled, multicenter trial, 13 patients with CLI and unsuitable for revascularization were assigned to receive intramuscular injection of placebo or FGF-4 using adenovirus as a vector (Ad5FGF-4).37 Dose-escalation of 2.87 x 108–1010 viral particle was administered into the ischemic muscle. Injections of Ad5FGF-4 were well tolerated and safe, and after 3-month follow-up, a trend toward more and slightly larger blood vessels was observed in the angiograms.37 However, the small sample size did not allow any firm conclusion regarding clinical efficacy.

Hypoxia-inducible factor-1α. In a phase I open label, double-blind, placebo-controlled trial, 34 patients with CLI Rutherford category 4 and 5 and unsuitable for revascularization were assigned to receive injections of placebo or HIF-1 into the ischemic leg muscle.38 HIF-1 was delivered in a dose-escalation of 1 x 108–2 x 1011 viral particles using a recombinant, replication-deficient adenovirus as a vector. No serious adverse events were attributable to study treatment and at one year, complete rest pain resolution was observed in 14 of 32 patients and complete ulcer healing in 5 of 18 patients. No amputations occurred in the two highest-dose groups.38

Hepatocyte growth factor. In a phase I/II double-blind, placebo-controlled, multicenter HGF-STAT trial conducted in the USA, 104 patients with rest pain or ischemic ulcers being poor candidates for standard revascularization were randomized to receive intramuscular injections of placebo or HGF plasmid into the ischemic leg muscle.39 HGF was delivered as follows: 0.4 mg at days 0, 14, and 28 (low dose); 4.0 mg at days 0 and 28 (middle dose); or 4.0 mg at days 0, 14, and 28 (high dose). Adverse events occurred in 86% of the patients and were not different between groups, and after 6-month follow up, the primary endpoint of TcPO2 increased in the high dose group as compared with placebo (24.0, 95% CI 15.5 to 32.4 mm Hg vs. 9.4, 95% CI 0.9 to 17.8; ANCOVA p = 0.0015). There was no difference between groups regarding secondary endpoints, including ABI, toe-brachial index, pain relief, wound healing, or major amputation.39

The recently completed phase II double-blind, randomized, placebo-controlled trial to assess the safety and efficacy of HGF plasmid to improve perfusion in CLI subjects who have peripheral ischemic ulcers, enrolled 27 patients. This study is sponsored by AnGes (AnGes, Inc, Gaithersburg, Maryland) and results are expected to be announced this year.40 These results provide the basis for a phase III randomized placebo-controlled trial, which will begin soon.

In another single-center prospective open-labeled study, 6 patients with CLI due to atherosclerosis and Buerger’s disease who were unsuitable for revascularization were treated with injections of HGF plasmid into the ischemic leg muscle.41 After a test dose of 0.4 mg intramuscular injections revealed no acute or subacute allergic reaction, a therapeutic dose of 2 mg of naked HGF plasmid DNA were administered into the ischemic muscle at 4 injection sites, and a second injection was similarly administered four weeks thereafter, giving a total dose of 4 mg plasmid DNA. No severe adverse effects (SAE) were attributable to gene transfer, and after 3-month follow up, a reduction of pain scale was observed in 5 of 6 patients, ABI was improved in all 5 patients, and the ischemic ulcer size was reduced more than 25%.41 These results supported the performance of an ongoing randomized, double-blind, placebo-controlled multicenter trial in patients with CLI in Japan using intramuscular injection of 4 mg HGF plasmid, which is repeated at 4 weeks.

Similar results were observed in a phase I dose escalation trial in which 12 patients with CLI not eligible for revascularization were randomly assigned to increasing doses of 2 mg to 16 mg of a novel, non-viral plasmid encoding 2 isoforms of HGF (pCK-HGF-X7).42 The two isoforms have a high level of gene expression and can induce the formation of collateral vessels through the activation of c-met pathway of the target cells, such as endothelial and smooth muscle cells. HGF was well tolerated with no serious adverse effects, amputation, or deaths through 1 year of follow up, and pain severity score was significantly reduced.42

In summary, early clinical trials have demonstrated encouraging results that demonstrate improved ulcer healing, increased vascularity, and perfusion and pain reduction. An overview of angiogenesis studies in CLI using these four angiogenic factors are shown by year of publication and sample size (Figure 2).

Safety Aspects of Gene Therapy

The inclusion criteria for these trials have generally included Rutherford class 4 and 5 (rest pain or nonhealing ulcers) with documented abnormal ankle or toe pressure or TcPO2 who were not amenable (or suboptimal) for standard surgical or percutaneous revascularization. In general, patients were excluded if they had undergone a successful revascularization procedure within the past 3 months, or had a history of cancer within the last 5 years, severe retinopathy, or macular degeneration. In addition, patients are required to have routine cancer screening, as recommended by the American Cancer Society, and ophthalmologic evaluations. Other more standard exclusion criteria include chronic hemodialysis, immunosuppressive medication, chronic inflammatory disease, bleeding disorders, imminent major amputation, women of child-bearing potential, hepatic disease, HIV-positivity, or life expectancy of less than 1 year.9,37,39,41 Therefore, these have been high-risk CLI patients who have limited treatment options, but who have been screened to eliminate potential adverse effects.

Results from numerous randomized, controlled trials suggest that gene transfer is feasible and safe with no serious adverse effects apparent at this time. Adverse effects have generally been consistent with the baseline rate for the population studied.18 However, long-term safety data from large scale trials is lacking.

Potential adverse effects include angiogenic stimulation of unrecognized malignancies, progression of diabetic retinopathy, macular degeneration, or progression of atherosclerosis due to angiogenic effects on the vasa vasorum. At present, however, experimental and clinical experiences with different growth factors have not identified an increased risk for malignancies, retinopathy, or acute coronary syndromes.18 Other less SAE have been reported, which included hypotension, vascular leakage, and transient lower limb edema when FGF or VEGF is used.18,43 A transient increase in C-reactive protein, proteinuria, and thrombocytopenia and renal insufficiency has also been reported in other angiogenic trials.21,33,44,45 HGF in angiogenic trials showed no greater adverse effects than the baseline rate of the population studied, which makes this agent very interesting. Overall, the initial trials demonstrated an excellent safety profile. However, ongoing systematic surveillance for safety is clearly needed in this expanding research field.

Future Perspectives

Gene therapy has slowly continued to develop as a potential treatment over the past two decades and has provided important knowledge of different growth factors used, selection of genes and vectors and the methods of administration. Numerous angiogenic phase I and II trials have been published showing minimal toxicity and substantial clinical improvement demonstrated in patients with CLI. Large-scale phase III trials, such as the multicenter double-blind, placebo-controlled trial evaluating efficacy and safety of NV1FGF in CLI patients with skin lesions (TAMARIS),46 are underway, which will include 490 patients assigned to placebo or intramuscular injection of NV1FGF. This study is sponsored by Sanofi aventis (Sanofi-aventis, Paris, France) and is expected to be completed in July 2010.

Another promising approach of therapeutic angiogenesis in patients with CLI unsuitable for revascularization is the use of cell-based therapy. There have been a number of early phase I clinical trials and several larger phase II randomized placebo-controlled trials are underway.47

Overall, despite the promising results of therapeutic angiogenesis in initial clinical trials, a large number of questions remain unanswered, including the optimal dose, dosing schedule, and route of administration. Given the complexity of angiogenesis, it is unclear which growth factor will have the most effective neovascularization potential and whether a combination of growth factors might be more effective.

Although these challenges remain, we believe that the ongoing trials will enhance our understanding of angiogenic mechanisms and might yet lead to the routine clinical use of therapeutic angiogenesis. Our understanding of angiogenic mechanisms will undoubtedly become more sophisticated, and new treatment strategies may use combinations of growth factors and/or cell therapy using endothelial progenitor cell. This approach may enhance angiogenesis in a complementary or synergistic manner.

Conclusion

Both plasmid DNA and adenoviral vectors have successfully been used to deliver a variety of angiogenic growth factors in patients with CLI to promote neovascularization. Clinical phase I/II studies have shown excellent safety and promising efficacy in surrogate endpoints, such as pain, ulcer healing, and perfusion imaging. Although many questions remain unanswered, therapeutic angiogenesis has the potential to revolutionize our approach to patients with CLI.

Acknowldegement. Dr. Keo has received a grant from Swiss National Science Foundation (PBBRB-121067).

___________________________

From 1University of Bern, Bern, Switzerland, 2Minneapolis Heart Institute Foundation at Abbott Northwestern Hospital, University of Minnesota School of Public Health, Minneapolis, Minnesota, 3Asklepios Klinik St, Georg, Germany.

Dr. Henry has received speaker honoraria from Sanofi Aventis and research grants from ViaMed.