Cardiovascular atherosclerotic diseases remain leading causes of morbidity and mortality in the world. Despite the significant progress that has been made in the management of these diseases using medical, surgical and percutaneous therapies over the last three decades, there remains a significant population of patients who are not optimal candidates for surgical or percutaneous revascularization.
Substantial research has focused on the administration of angiogenic growth factors, either as recombinant protein or by gene transfer, to promote the development of supplemental collateral blood vessels that will constitute endogenous bypass conduits around occluded native arteries; a strategy termed "therapeutic angiogenesis". In large numbers of pre-clinical trials of ischemia this strategy has demonstrated augmentation of tissue perfusion through neovascularization. In patients with critical limb ischemia or end-stage coronary artery disease, clinical trials have shown symptomatic improvement and objective evidence of improved perfusion, suggesting that this strategy may constitute an alternative means of treatment for patients in whom contemporary therapies have previously failed or are not feasible.
Future goals of angiogenesis research will be to determine: the optimal dose, formulation, route of administration and combinations of growth factors; the requirement for endothelial progenitor cell or stem cell supplementation; to provide effective and safe therapeutic angiogenesis; as well as tailoring angiogenesis to individual patient needs. This review will discuss the biology of neovascularization as related to angiogeneic growth factors and then review gene transfer strategies for critical limb and myocardial ischemia.
NEOVASCULARIZATION
The strategies of gene transfer for therapeutic angiogenesis are ultimately based upon the findings of Folkman and colleagues1 who suggested that the establishment and maintenance of a vascular supply is an absolute requirement for growth of normal as well as neoplastic tissue as a result of two main processes, vasculogenesis and angiogenesis, and that such neovascularization involves angiogenic growth factors. Vasculogenesis is the de novo in-situ differentiation of endothelial cells (ECs) from mesodermal precursors in the embryo by association of endothelial progenitor cells (EPCs) or angioblasts, and their subsequent reorganization into a primary capillary plexus2.
Angiogenesis is the formation of new blood vessels from preexisting blood vessels induced by the proliferation and migration of pre-existing, fully differentiated ECs resident within parent vessels in response to stimuli such as hypoxia, ischemia, mechanical stretch, and inflammation3,4. Angiogenesis can be both physiological and pathological, the former being processes such as wound healing and the female reproductive cycle, the latter processes involving angiogenesis include tumor growth, rheumatoid arthritis and proliferative diabetic retinopathy.
Vasculogenesis was previously considered restricted to embryonic development while angiogenesis, although recognized to also occur in the embryo, was thought to be exclusively responsible for postnatal neovascularization. More recent evidence suggests that the basis for native, as well as therapeutic neovascularization, is not restricted to angiogenesis, but encompasses both embryonic processes. Circulating CD34 antigen-positive EPCs were recently isolated from adult species and shown to differentiate along an endothelial cell lineage in vitro5 thus constituting inferential evidence for circulating stem cells. In addition, the demonstration that bone marrow-derived EPCs are increased in number in response to tissue ischemia6, home to and incorporate into foci of neovascularization in adult animals7, can augment collateral development following ex vivo expansion and transplantation8 suggests that neovascularization in the adult is not restricted to angiogenesis but involves "postnatal vasculogenesis".
Angiogenic growth factor-induced neovascularization encompasses a range of vessel caliber, from medium-sized arteries visualized by angiography to increased capillary density demonstrated by post-mortem histology. A proportion of newly recognized medium-sized arteries may develop as a result of "arteriogenesis" or in-situ proliferation of pre-existing arteriolar connections into larger collateral vessels9 by remodeling; it is unknown whether such remodeling occurs as a direct result of growth factor modulation or as a flow-mediated maturation of these collateral conduits.
ANGIOGENIC GROWTH FACTORS
While many cytokines have angiogenic activity, the best studied both in animal models and clinical trials are vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF).
Vascular Endothelial Growth Factor
The human VEGF genes that have been identified to date are VEGF-1, VEGF-2 or VEGF-C, VEGF-3 or VEGF-B, VEGF-D, VEGF-E, and placental growth factor (PIGF). All are encoded by different genes and are localized to different chromosomes, but share considerable homology. The principal cellular target of VEGF is the EC. There are three known endothelial-specific fms-like tyrosine kinases VEGFR-1 (Flt-1), VEGFR-2 (Flk-1/KDR) and VEGFR-3. Hypoxia induces the formation of VEGF by the ECs and leads to up-regulation of the VEGF receptors10. VEGFR-1 generates signals that organize the assembly of ECs into tubes and functional vessels11. VEGFR-2 is responsible for EC proliferation and migration12,13. VEGFR-3 (Flt-4) principally mediates lymphangiogenesis14.
VEGF possesses several features that facilitate gene transfer. First, VEGF contains a secretory signal sequence that permits the protein to be secreted naturally from intact cells thus enabling a sequence of additional paracrine effects to be activated15. Second, its high-affinity binding sites are exclusive to ECs; consequently, the mitogenic effects of VEGF are limited to ECs, in contrast to acidic and basic FGF, both of which are known to be mitogenic for smooth muscle cells and fibroblasts as well as ECs16,17. Third, VEGF possesses an autocrine loop that shared by most angiogenic cytokines and facilitates modulation of EC behavior; when activated under hypoxic conditions, the autocrine loop serves to amplify and thereby protract the response in ECs stimulated by exogenously administered VEGF. Furthermore, factors secreted by hypoxic myocytes upregulate VEGF receptor expression on ECs within the hypoxic milieu. Such localized receptor expression may explain the finding that angiogenesis does not occur indiscriminately, but rather at sites of tissue ischemia. An important additional role for VEGF is augmentation of circulating EPC numbers, documented in mice and humans following VEGF gene transfer18-20.
Fibroblast Growth Factor
FGF is a family of nine factors namely basic FGF, acidic FGF, FGF 3-9. Acidic FGF [aFGF or FGF-1] and basic FGF [bFGF or FGF-2] are the most extensively characterized members of the large FGF family. FGF’s are non-secreted growth factors lacking a signal peptide sequence. Extracellular release of FGF is caused by cell death or damage. Although FGFs are potent EC mitogens, they are not EC specific, and also serve as ligands for other cell types including vascular smooth muscle cells and fibroblasts. At least 4 high-affinity FGF receptors have been identified and their cDNAs have been cloned. The FGFs, like VEGF, also stimulate EC synthesis of proteases including plasminogen activator and metalloproteinases, important for extracellular matrix digestion in the process of angiogenesis21. Unlike VEGF, however, the common forms of FGF (FGF-1 and 2) lack a secretory signal sequence; clinical trials of FGF gene transfer have consequently required either modification of the FGF gene22 or use of another of the FGF gene family with a signal sequence23,24.
THERAPEUTIC ANGIOGENESIS
Angiogenic cytokines may be administered as recombinant protein or as genes encoding for these proteins. Given that both protein and gene delivery approaches have been relatively well tolerated thus far in clinical trials, ongoing investigations will determine the optimal preparation and delivery strategy for therapeutic neovascularization.
Protein therapy remains the more conventional approach and some investigators have indicated that this strategy is the closest to practical use. Nevertheless, recombinant protein is usually administered systemically, and is therefore limited by potential adverse effects of the high plasma concentrations required to achieve adequate tissue uptake.
Gene transfer is introduction of genetic material into somatic cells of an organism with the aim of achieving high levels of sustained gene expression without provoking adverse host reactions. The success of gene transfer strategies depends on the efficiency with which the transgene is introduced and expressed into the target cell and the duration of transgene expression. Transfer vectors facilitate cellular penetration and intracellular trafficking of the transgene and local delivery systems deliver the vector to the vicinity of the target cells. There are two major categories of gene transfer systems, viral and nonviral. The most commonly used viral vectors for gene transfer are adenovirus and retrovirus. The nonviral methods include introduction of naked DNA into the target area and the use of liposomes.
Ischemic muscle represents a promising target for gene therapy. Striated and cardiac muscles have been shown to take up and express naked plasmid DNA as well as transgenes incorporated into viral vectors. Moreover, previous studies have shown that the transfection efficiency of IM gene transfer is augmented more than five-fold when the injected muscle is ischemic25,26. However, while viral vectors may enhance transfection efficiency and thus yield higher levels of gene expression, in vitro27 and in vivo28 models demonstrated that low-efficiency, but site-specific, transfection (successful transfection in <1 of cells with a gene plasmid dna encoding for secreted protein e g vegf may overcome the handicap inefficient transfection by secreting adequate to achieve local levels physiologically meaningful biological effects therefore achieving therapeutic not realized genes proteins that remain intracellular bfgf furthermore unlike viral vectors does induce inflammation
CRITICAL LIMB ISCHEMIA
In a large proportion of patients with critical limb ischemia, the distribution and extent of the arterial occlusive disease makes percutaneous or surgical revascularization impossible. The consensus statement of the European Working Group on Critical Limb Ischemia29 states that no medical treatment has been shown to alter the natural history of critical limb ischemia. In addition, quality-of-life indices for these patients are similar to those patients with terminal stages of malignancy. Despite the associated morbidity and mortality associated with amputation, it is often chosen as first-line therapy. Consequently, the need for alternative treatment strategies in patients with critical limb ischemia is compelling. Significant research has focused on developing angiogeneic therapies to and thereby provide novel approaches to the treatment of limb ischemia.
VEGF Gene Transfer in Peripheral Ischemia
Evidence that VEGF stimulates angiogenesis in vivo had been developed in experiments performed on rat and rabbit cornea, the chorioallantoic membrane and the rabbit bone graft model16,30. Preclinical studies have established proof of principle for the concept that the angiogenic activity of VEGF is sufficiently potent to achieve therapeutic benefit; augmentation of angiographically visible collateral vessels and histologically identifiable capillaries were demonstrated in rabbits with severe, unilateral hindlimb ischemia31,32. Subsequently, angiographic and histologic evidence of angiogenesis was subsequently demonstrated following intra-arterial gene transfer of phVEGF165 in a human patient33.
However, intra-arterial delivery has several inherent limitations that could undermine successful gene transfer for critical limb ischemia. In the case of naked DNA i.e. DNA unassociated with viral or other adjunctive vectors, cellular uptake is virtually nil when the transgene is directly injected into the arterial lumen, presumably due to prompt degradation by circulating nucleases. In addition, the diffuse distribution of neointimal thickening and/or extensive calcific deposits may limit gene transfer to the smooth muscle cells of the arterial media34.
Pre-clinical studies were therefore designed to establish the feasibility of site-specific IM gene transfer of VEGF in critical limb ischemia to promote therapeutic angiogenesis. Meaningful biological outcomes were observed following VEGF gene transfer of naked DNA by direct injection into skeletal muscle of ischemic rabbit hindlimbs26,35 as evidenced by increased hindlimb blood pressure ratio, increased Doppler-derived iliac flow, enhanced neovascularity by angiography and increased capillary density at necropsy.
That intramuscular VEGF gene transfer might be utilized to successfully accomplish therapeutic angiogenesis in patients with critical limb ischemia36 was demonstrated in patients with non-healing ischemic ulcers and/or rest pain by intramuscular gene transfer of 4000 µg naked plasmid DNA encoding VEGF (phVEGF165). Gene expression was documented by a transient increase in serum levels of VEGF monitored by ELISA. Therapeutic benefit was demonstrated by regression of rest pain and/or improved limb integrity, increased pain-free walking time and ankle-brachial index, newly visible collateral vessels by digital subtraction angiography and qualitative evidence of improved distal flow by magnetic resonance imaging.
Subsequent clinical trials with phVEGF165 have utilized randomized, (blinded) IM injections in 55 patients (ages 24-84 yrs, m=56.7yrs) with ischemic rest pain (n=14) or ischemic ulcers (n=41). Evidence of clinical improvement was observed in 13/14 (72%) patients with rest pain alone and 26/41 (63%) patients with ischemic ulcers over a follow-up period of 4-36 months. For the total cohort of 55 patients, a favorable clinical outcome was achieved in 65.5%. Multiple logistic regression analysis identified rest pain and age < 50 years as significant (p≤0.05) predictors of a favorable clinical outcome. Diabetes, smoking, hyperlipidemia, hypertension and phVEGF165 dose were not predictors of clinical outcome37. Complications in these patients have been limited to lower-extremity edema that develops in approximately one-third of patients38.
A similar treatment strategy was used in 11 patients with Buerger’s disease presenting with critical limb ischemia, nine of which were successfully treated39. These patients had resolution of nocturnal rest pain and healing of foot and/or leg ulcers. The ankle-brachial index increased by greater than 0.1 and newly formed collateral vessels were seen on MRA and serial contrast angiography.
Pre-clinical studies from our laboratory demonstrated that VEGF-2 could promote angiogenesis in a rabbit hind-limb ischemia model40 and stimulate the release of nitric oxide from ECs40. Subsequently randomized, double-blinded, placebo-controlled, dose-escalating trials have commenced to investigate the therapeutic potential of VEGF-2 gene transfer in patients with critical limb ischemia (CLI). A total of 46 patients with CLI have thus far been randomized on a 3:1 (treatment:placebo) basis to receive saline or pVEGF2 as naked DNA injected directly into ischemic lower extremity muscles; of 46 patients, 21 had rest pain alone, and 25 had ischemic ulcers ± rest pain. The results of this phase I trial are currently pending.
Other Strategies for Patients with Peripheral Arterial Disease
The potential for basic fibroblast growth factor (bFGF) to improve collateral development in animal models of hind limb ischemia has been previously demonstrated41-43. The safety of intra-arterial bFGF administration in patients with intermittent claudication was recently demonstrated by Lazarous et al44. In this phase I, double blind, placebo-controlled clinical trial there was improvement in calf blood flow by strain gauge plethysmography in bFGF-treated patients at 6 months compared to controls. However, further larger scale trials are warranted to investigate the safety and efficacy of FGF in patients with peripheral arterial disease, particularly those with critical limb ischemia.
MYOCARDIAL ISCHEMIA
For patients in whom anti-anginal medications fail to provide sufficient symptomatic relief, other interventions such as angioplasty or bypass surgery may be required. While both types of intervention have been shown to be effective for various types of patients, a considerable group of patients may not be candidates for either intervention due to the diffuse nature of their coronary artery disease. Moreover, there are many patients in whom recurrent narrowing and/or occlusion of bypass conduits after initially successful surgery has left the patient again symptomatic with no further option for conventional revascularization.
For the purposes of myocardial angiogenesis, angiogenic cytokines have been administered via a wide variety of routes that include intravenous, intracoronary, transepicardial at time of bypass surgery or via thoracotomy, intra-pericardial or peri-adventitial at time of bypass surgery and most recently transendocardial by catheter. Clinical trials have thus far favored the intracoronary (adenovirus) or direct myocardial (naked DNA or adenovirus) route because of the belief that local delivery of recombinant protein or gene is considered ideal45. Future progress mandates that the inherent risk of an operative approach be circumvented with the percutaneous catheter-based approach.
Pre-Clinical Studies with VEGF
Following demonstration of proof of principle that cytokine gene transfer could be used to promote angiogenesis in humans with CLI, we extrapolated this strategy to myocardial ischemia. While animal experiments performed in our laboratory utilizing recombinant human VEGF (rhVEGF165) protein administered directly into the left coronary ostium demonstrated significant augmentation of flow to collateral-dependent ischemic myocardium this was complicated by hypotension, apparently mediated by VEGF-induced release of NO46. Similar results were reported from other groups utilizing intracoronary injection in the pig47 and dog48. Intramyocardial49 and peri-adventitial47,50 administrations of VEGF protein showed limited efficacy but intravenous infusion was ineffective49.
Accordingly, we anticipated that local expression of VEGF for a protracted period of 2-3 weeks might circumvent the problem of symptomatic hypotension, yet still achiev e a reduction in myocardial ischemia. Consequently we demonstrated safe and successful direct myocardial gene transfer of phVEGF16551,52 or VEGF-253 via a minimally invasive chest wall incision in a swine model of chronic myocardial ischemia with enhanced collateral vessel filling and improved perfusion to ischemic myocardium by colored microspheres.
Intramyocardial injection of adenovirus encoding VEGF12154,55 via thoracotomy in a pig ameroid model improved collateral perfusion and function. Intracoronary adenoviral gene delivery produced much lower gene and VEGF levels in the myocardium with poor localization55. Pericardial delivery of adenovirus encoding VEGF165 in a dog model did not increase collateral flow56.
Recent studies have suggested that catheter-based myocardial gene transfer of naked plasmid VEGF165 and VEGF-2 is effective in the pig51. This less invasive approach to intramyocardial gene transfer has been shown to achieve suitable gene expression51,57-59.
Pre-Clinical Studies Utilizing FGF
A series of animal experiments have demonstrated that intracoronary FGF-2 improves myocardial perfusion and function and increases collateral flow in the dog in both acute and chronic ischemia41,60-63. This improvement was as a result of angiographically documented and histologically identifiable sprouting of capillary network from the original coronary vessels. Beneficial effects on collateral flow and left ventricular function were also seen with FGF-2 in the pig after single doses administered perivascularly or into the pericardium64-66. Recombinant FGF-1 protein was ineffective in the dog67,68. Experience with FGF gene transfer is more limited. Single-dose intramuscular injection of naked DNA encoding FGF-1 and intracoronary adenoviral transfer of the FGF-5 gene have each been shown to improve flow in the rabbit hindlimb22 and porcine myocardium23 respectively.
>Clinical Trials of Direct Myocardial VEGF Gene Transfer
Published studies of VEGF gene transfer for therapeutic angiogenesis in human subjects have thus far been limited to Phase I dose-escalating, non-randomized trials involving naked plasmid DNA and adenovirus assisted. Patients in these trials generally have severe angina refractory to medical therapy, demonstrate ongoing myocardial ischemia and are unsuitable for conventional revascularization.
As a result of the above animal experiments of VEGF plasmid DNA gene transfer, our center initiated a phase 1, dose escalating, open label clinical study to determine the safety and bioactivity of direct myocardial gene transfer of phVEGF165 as sole therapy (i.e. without angioplasty, stenting or bypass graft surgery) for patients with stable exertional angina refractory to medical therapy, areas of viable but under-perfused myocardium on perfusion scanning and multi-vessel occlusive coronary artery disease. Preliminary results of this trial suggested that safe and successful transfection could be achieved by this method with a favorable clinical effect69,70.
30 patients received phVEGF165 administered by direct myocardial injection in 4 aliquots of 2.0 ml via a "mini-thoracotomy"; total dose 125µg (n=10), 250µg (n=10), 500µg (n=10). An immobile field for intramyocardial injection was ensured by using a stabilizing device that facilitates vascular anastomosis during beating heart bypass. Continuous transesophageal echocardiographic monitoring was performed throughout the procedure to monitor development of wall motion abnormalities associated with injections and ensure that plasmid DNA was not injected into the LV cavity71. No perioperative complications were experienced. There was no evidence of myocardial damage by cardiac enzyme analysis and patients maintained left ventricular function. Gene expression was documented by a transient but significant increase in plasma levels of VEGF monitored by ELISA assay. All patients experienced marked symptomatic improvement and/or objective evidence of improved myocardial perfusion. At latest follow-up, 15/30 patients at 360 days were free of angina. Specifically, sublingual nitrate use fell from 60/week to 3/week at day 360 accompanied by a significant reduction in episodes of angina from 56/week to 4/week at day 360. Exercise time for the group at 360 days had increased by 98 seconds and exercise time to angina increased by 2.5 minutes over baseline. There were two late deaths (4.570 and 28.5months), and one patient underwent a cardiac transplant at 13 months.
Evidence of reduced ischemia on SPECT-sestamibi myocardial perfusion scanning was documented in 22/29 patients with a significant reduction in both stress and rest mean perfusion/ischemia score at day 60 follow-up. It is intriguing to note that not only defects observed in the perfusion scans with pharmacologic stress, but also those observed at rest, improved post-gene transfer; sequential SPECT scans recorded before and after gene transfer demonstrated partial or complete resolution of fixed defects in 4 (33%) and 5 (43%) patients respectively in whom defects were present on the initial rest image. This is consistent with the notion that these pre-existing defects constitute foci of hibernating viable myocardium72-74that have resumed or improved contractile activity as a result of therapeutic neovascularization. This observation was supported by the findings of electromechanical mapping utilized in the final 13 consecutive patients. Resting perfusion defects on the SPECT images corresponded to areas with ischemic characteristics (reduced wall motion with preserved viability) on the endocardial maps. Foci of ischemia were identified pre-operatively in all patients, with significant improvement in these endocardial wall motion abnormalities at 60 days post gene transfer75.
This study provides the first evidence for a favorable clinical effect of direct myocardial injection of naked plasmid DNA encoding for VEGF as the sole therapeutic intervention. Similar favorable experience has been realized in an open-label, dose-escalating, multicenter clinical trial of VEGF-2 plasmid DNA in 30 patients with end-stage coronary artery disease and refractory class III or IV angina. In all patients, there were no procedural adverse events, although there was one death, 20 hours after surgery. At 12 months following gene transfer, the mean number of anginal episodes and nitrate table consumed per week decreased significantly, 25/29 patients or 86% had improved by 2 or more angina classes and mean duration of exercise increased by more than 2 minutes (unpublished data).
The only other reported study of direct myocardial VEGF gene transfer was with adenoviral-assisted VEGF121 injection to patients undergoing bypass graft surgery [n=15], and as sole therapy via mini-thoracotomy [n=6]. Symptoms and exercise duration improved in both bypass surgery and sole therapy groups but stress-induced nuclear perfusion images remained unchanged. The data in this study is consistent with the concept that adenovirus VEGF121 appears to be well tolerated in patients with advanced coronary disease.
This early experience with myocardial VEGF gene transfer, while encouraging from the standpoint of therapeutic angiogenesis and gene therapy, leaves multiple issues unresolved. Optimizing the anatomic site, number, and dose of direct myocardial injections will require further investigation. The strategy of gene therapy alone administered via a mini-thoracotomy does not permit randomization against placebo (untreated controls) or clinical testing of alternative dosing regimens including multiple treatments.
Catheter-Based Myocardial VEGF Gene Transfer
While successful intravascular33, pericardial76 and intramuscular36 gene transfer have all been performed using minimally invasive delivery techniques, all of the aforementioned work involving myocardial gene transfer has to date required an operative procedure.
Preliminary pre-clinical studies utilized a previously described navigation system and catheter mapping technology (NOGA™) integrated with an injection catheter (Biosense-Webster, Warren, NJ), the distal tip of which incorporates a 27G needle to deliver six injections (1.0ml per injection) to the myocardium of normal and ischemic swine57 to determine the safety and feasibility of catheter-based gene transfer. Results with methylene blue suggested safe, reliable and reproducible targeting of endocardial sites and injection of a reporter gene (pCMV-nls LacZ ) demonstrated peak ß-galactosidase(ß-gal) activity (greater in ischemic versus non ischemic myocardium indicating enhanced gene transfer in ischemic myocardium) in the target area with low level to negligible activity seen in areas remote from the injection sites suggesting relatively localized gene transfer. Similar findings were demonstrated by a study utilizing adenoviral-assisted gene transfer of a reporter gene77. These results established that percutaneous myocardial gene transfer could be successfully achieved in normal and ischemic myocardium in a relatively site-specific fashion without significant morbidity or mortality. The mapping capabilities of the NOGA™ system utilized in this study were useful for demonstrating that gene expression could be directed to pre-determined LV sites, indicating that this technique clearly may be advantageous for avoiding gene transfer to sites of myocardial scar as well as to accurately relocate the tip of an injection catheter toocardial ischemia (or hibernating myocardium) where gene transfer may be potentially optimized.
Preclinical studies were also performed to specifically test the feasibility and safety of catheter-based delivery of naked plasmid DNA encoding for VEGF-1 and VEGF-251. Effective GTx was demonstrated by presence of plasmid DNA in myocardial tissue by PCR. No VEGF protein or plasmid was identified in remote organs. Injections caused no hemodynamic changes, no sustained ventricular arrhythmias and there was no ECG evidence of infarction. Objective evidence of reduced ischemia (reduced ischemic area on NOGA™ mapping) was documented in all VEGF-transfected animals. No improvement was seen in the control animals. These findings therefore suggested that percutaneous myocardial injection of VEGF could be safely and reproducibly accomplished in ischemic myocardium of swine.
Subsequently, we initiated a pilot study of percutaneous, catheter-based VEGF-2 DNA gene transfer or a sham procedure guided by the NOGA™ mapping system in 6 patients with non-revascularizable symptomatic myocardial78. VEGF2-transfected patients reported significant reduction in weekly anginal episodes and nitrate tablet consumption at 12 months post gene transfer. In contrast, while blinded patients randomized to the control group reported an initial reduction in these parameters, this changed clinical profile was not sustained past 30 days, suggesting that the continued reduction in angina in the VEGF-2treated group was not a placebo effect. The symptomatic improvement was again accompanied by objective evidence of improved myocardial perfusion by both SPECT-sestamibi perfusion scanning and electromechanical mapping78. While the clinical findings of this pilot trial concerning efficacy are similarly encouraging, the number of patients and the single-blinded design preclude firm conclusions in this regard. Consequently, a multi-center randomized, double blind, placebo-controlled trial of cathetthat has thus far enrolled 19 patients. There have been no complications associated with a total of 1eit.
This preliminary experience thus suggests that it is feasible to replace currently employed operative approaches with minimally invasive techniques for applications of cardiovascular gene therapy designed to target myocardial function and perfusion. Such an approach may have at least three advantages compared to an operative approach. First, it potentially allows more selective delivery of the transgene to targeted ischemic zones, including sites that are less accessible by a mini-thoracotomy. Second, the catheter-based approach, because it obviates the need for general anesthesia and operative dissection through adhesions related to placement of previous bypass conduits, facilitates placebo-controlled, double-blind testing of myocardial GTx. Third, the intervention can be performed as an out-patient procedure, and repeated if necessary.
Myocardial Gene Transfer with FGF
Clinical experience with FGF gene transfer for myocardial angiogenesis is limited. Perivascular FGF-2 (bFGF) or placebo contained within heparin alginate microcapsules were implanted in the subepicardial fat in a non-graftable myocardial territory in patients undergoing CABG with viable and ischemic myocardium79. In this study, there was a relatively high peri-operative mortality and morbidity, attributed to the advanced nature of coronary artery disease. However, preliminary data suggested improvement in angina, myocardial perfusion and regional function by MRI in the bFGF group.
Clinical Studies of Myocardial Recombinant Protein Therapy
The first clinical study of recombinant protein for myocardial ischemia employed intramyocardial injections of FGF-1 (aFGF) in conjunction with CABG in patients undergoing left internal mammary (LIMA) bypass of the left anterior descending (LAD) coronary artery. In this randomized, double-blinded, placebo controlled study of 40 patients (20 in FGF group and 20 in placebo group), angiography revealed that FGF-treated patients group had an increased contrast uptake (sprouting capillary network) at the site of growth factor injection compared to the placebo group. This effect was supported by improved functional class and reduction in nitrate consumption at 3 year follow-up80. The same group also recently showed improved SPECT perfusion and exercise capacity at 12 weeks in patients with severe angina unsuitable for conventional revascularization injected transepicardially via mini-thoracotomy with rhFGF-181.
The recently reported VIVA trial, a double-blind, placebo-controlled, dose escalating trial of patients with viable myocardium who were not optimal candidates for percutaneous or surgical revascularization compared 2 doses of VEGF-1 protein to placebo in 178 patients given a single intracoronary infusion followed by 3 separate intravenous infusions82-85. Doses were limited by hypotension that developed at higher doses in an earlier dose-ranging study. Improvement in exercise treadmill duration (approximately 45 seconds) was similar in treatment and placebo groups at 60 days. At 120 days, the high dose group maintained the improvement with an increase of 47 seconds over baseline, while the placebo group showed only a 14-second improvement. There were no significant differences from placebo in angina grade or quality of life measures at 60 days though there was a significant reduction in angina grade at 120 days in the high dose group. Angiographic and SPECT sestambi scan did not show any significant change in any group.
The FIRST study compared a single intracoronary dose of recombinant FGF-2 (bFGF) with placebo in 337 patients, delivered as a single 20 min infusion divided between two major sources of coronary blood supply in patients with non revascularizable coronary artery disease86. The 90 day results failed to show significant differences from placebo in the primary endpoint of exercise time (65 vs 45 seconds improvement, p=0.64), or in rest or stress nuclear perfusion.
ARTERIAL GENE THERAPY FOR INHIBITING RESTENOSIS IN PATIENTS WITH CLAUDICATION UNDERGOING SUPERFICIAL FEMORAL ARTERY ANGIOPLASTY
Superficial femoral artery (SFA) stenosis represents one of the most common sites of peripheral vascular obstruction. Percutaneous transluminal angioplasty (PTA) has been used widely and successfully to treat atherosclerotic obstructions in the peripheral and coronary circulations. However, restenosis following angioplasty of the SFA/popliteal artery continues to be a vexing, and consequently, expensive complication of this otherwise efficacious intervention. While acute procedural success for percutaneous revascularization of lesions in the SFA using conventional guidewires and standard PTA is well in excess of 90%, published reports have established that restenosis may complicate the clinical course of as many as 60% of patients undergoing PTA for SFA stenosis and/or occlusion. Previous strategies to limit the development of restenosis by non-mechanical means have not proved effective. Treatment strategies aimed at specifically restoring endothelial integrity have not been previously explored for restenosis prevention. Animal studies demonstrated that administration of mitogens, such as VEGF, that promote EC migration and/or proliferation might achieve acceleration of re-endothelialization and thereby reduce intimal thickening87-90.
We therefore designed a phase I, single site, dose-escalating open-label, unblinded gene therapy trial to accelerate re-endothelialization at the site of PTA-induced endothelial disruption as a novel means to inhibit restenosis following PTA. The primary objective of this study was to document the safety of percutaneous catheter-based delivery of the gene encoding VEGF, in patients with claudication due to SFA obstruction.
Arterial VEGF gene transfer has thus far been performed in 20 patients, 13 males and 7 females with a mean age of 69 years. All patients had 2 or more cardiovascular risk factors. Gene expression was documented by a rise in plasma levels of VEGF. Peak plasma levels were recorded at a mean of 12 days following gene transfer. Mean claudication time increased from 2 minutes at baseline to 5 minutes up to 18 months post-gene transfer. Prior to gene transfer, all patients were classified as Rutherford Class 3. At 12-18 months following gene transfer, 10 patients were asymptomatic and 5 patients were class 1. After an initial improvement in 2 Rutherford classes following revascularization, 4 patients returned to class 3. One patient developed critical limb ischemia and required salvage therapy with intramuscular gene transfer of naked plasmid DNA encoding VEGF.
There was a significant and sustained improvement in ankle brachial index post gene transfer compared to baseline. The mean ABI was 0.70 pre gene transfer and was 0.89 up to 18 months post gene transfer. SFA stenosis in 15 patients dropped from a mean of 82% at baseline to 32% at an average of 9 months following gene transfer. These results were supported by IVUS findings at the time of follow-up angiography. Five patients had evidence of restenosis at angiography performed 6-12 months following gene transfer. Target vessel revascularization was required in all 5 patients. Histology from 3/4 patients to undergo directional atherectomy at the time of repeat revascularization for restenosis demonstrated active smooth muscle cell proliferation and high levels of proliferating cell nuclear antigen indicating extensive proliferative activity.
Thus, 20 patients have been treated with arterial VEGF gene transfer for prevention of restenosis. VEGF expression has been documented by ELISA assay. At 12-18 months follow-up, 5 out of 20 patients (25%) required target vessel revascularization for angiographic an d ultrasound evidence of restenosis. This preliminary study has suggested that gene therapy designed to accelerate re-endothelialization at the site of PTA-induced endothelial disruption can be safely performed. Importantly no evidence of accelerated atherosclerosis or an increase in the restenosis rate was observed following gene transfer.
POTENTIAL SAFETY CONCERNS
Many angiogenic factors are known to be involved in tumor growth secondary to enhancing angiogenesis which is vital for growth of most tumors, particularly solid tumors. Hence, in theory, angiogenic growth factors may lead to development of tumors and they may be too small to be recognized. Even so, there are neither in vitro nor in vivo data to suggest that VEGF increases the risk of neoplastic growth and/or metastases, although longer term follow-up will be required to address this issue in clinical trials. It was very interesting to note in VIVA Trial that there was a greater incidence of tumors in placebo group then in the VEGF group. This exemplifies the fact that the age group receiving such therapy will develop some unrelated tumors. Nevertheless, one must be vigilant about the possibility of cancer in patients treated with these angiogenic growth factors. In addition, concerns regarding the development of angiomata were raised in studies involving mice91 or rats92 treated with transduced myoblasts or supraphysiologic doses of plasmid DNA respectively. Importantly, no other preclinical or clinical reports, including those using adenoviral vectors, have described this complication.
It is theoretically possible that VEGF may exacerbate proliferative and/or hemorrhagic retinopathy in patients with diabetes in view of the high VEGF levels demonstrated in the ocular fluid of patients with active proliferative retinopathy leading to loss of vision93. To date, this adverse effect of therapeutic angiogenesis has not been observed. The local delivery of naked plasmid DNA encoding for VEGF-1 or VEGF-2 to more than 100 patients (one third with diabetes diabetes and/or remote retinopathy) treated at our institution with up to 4-year follow-up did not effect the visual acuity or fundoscopic findings as evidenced by serial funduscopic examinations pre- and post-gene transfer by an independent group of retinal specialists.
Experiments in transgenic mice engineered to overexpress VEGF ± angiopoietin have been demonstrated lethal permeability-enhancing effects of VEGF94. However, even though VEGF has been reported to cause local edema which manifest as pedal edema in patients treated with VEGF for critical limb ischemia, it responds well to treatment with diuretics38.
Therapies with recombinant proteins have been noted to produce hypotension46,95, particularly when used systemically and in higher doses; due to the fact that VEGF upregulates nitric oxide synthesis96,97; this complication, however, has never been described following gene transfer in either animals or humans.
Hematological side effects in the form of anemia and thrombocytopenia were reported following systemic bFGF administration in dogs. However, these adverse effects have not been reported in short courses of therapy. Renal toxicity in the form of proteinuria has also been report in animal studies of bFGF. This adverse effect is supposed to be transient and reversible98.
Another concern stems from the recent demonstration that inhibitors of angiogenesis tested in an apolipoprotein E deficient mouse model of atherosclerosis inhibited plaque growth and intimal neovascularization99. However, data available from four separate animal studies87-90 and two clinical studies of human subjects100,101 fail to support the notion that accelerated atherosclerosis is a likely consequence of administering angiogenic cytokines; the outcome, in fact, is quite the opposite, in that administration of VEGF led to a statistically significant reduction in intimal thickening due to accelerated re-endothelialization, thereby refuting the notion that acceleration of atherosclerosis will be a consequence of VEGF-induced stimulation of angiogenesis.
The current clinical strategies employed for critical limb and chronic myocardial ischemia constitute an extrapolation from initial applications of gene transfer to animal models with limb ischemia utilizing the 165-amino acid isoform of the VEGF-1 gene. These results, however, likely have generic implications for strategies of therapeutic neovascularization using alternative candidate genes, vectors, and delivery strategies. Preclinical data supporting the use of other VEGF-1 isoforms102 as well as other VEGF genes40, has been previously reported, as have preclinical studies using fibroblast growth factor (FGF)23,103; all of these are being actively studied in on-going clinical trials. Furthermore, the relative merits of gene transfer versus recombinant protein administration remain to be clarified.
The otherwise negative primary endpoint results of the VIVA and FIRST studies using intracoronary ± intravenous protein administration underscore the concern that the pharmacokinetics of recombinant protein administered into the vascular space may lead to inadequate local delivery of angiogenic growth factor within the ischemic myocardium. Clearly, additional investigations comparing doses of recombinant protein and routes of delivery will be required to resolve this issue. Until all of these studies are complete, the ideal method of achieving therapeutic angiogenesis remains unknown. In addition, results of Phase 1 studies, designed by definition to assess safety, must be interpreted with caution. Typically, the number of patients enrolled in such trials is relatively small, and for those lacking a control group, a placebo effect cannot be excluded. For studies in which recombinant protein or gene is administered in conjunction with conventional revascularization it may be difficult to determine the relative contributions of the angiogenic agent versus bypass surgery to the symptomatic response.
It is clear, however, that site-specific VEGF gene transfer can be used to achieve physiologically meaningful therapeutic modulation of vascular disorders and specifically that IM injection of naked plasmid DNA achieves constitutive overexpression of VEGF sufficient to induce therapeutic angiogenesis in selected patients with CLI. Of note, neither our intra-arterial animal studies, nor our human clinical experience utilizing naked plasmid DNA encoding for VEGF have disclosed any evidence of immunologic toxicity. Furthermore, at this early stage of clinical trials into myocardial gene therapy, it has been shown that direct myocardial gene transfer utilizing different doses of naked plasmid DNA encoding for VEGF165 and VEGF-2 can be performed safely and this approach augments myocardial perfusion. In terms of safety, no operative complications, and no aggravated deterioration in eyesight due to diabetic retinopathy104 have been observed in patients treated with phVEGF165 gene transfer. With specific regard to mortality, it should be noted that the cumulative mortality for the 85 patients with class 3 or 4 angina, all of whom were refused for conventional revascularization, undergoing operative or percutaneous naked DNA gene transfer of VEGF-1 or VEGF-2 has been 3/85 or 3.5% at up to 33 months follow-up. This compares favorably with an average 11-13% 1-year mortality for a similar group of almost 1000 patients receiving laser myocardial revascularization or continued medical therapy in 5 contemporary controlled studies105-109. On-going clinical studies will determine the potential for neovascularization gene therapy to be performed by non-surgical, catheter-based delivery, although early results are encouraging from a therapeutic standpoint.
For the most part, clinical studies of therapeutic angiogenesis have been restricted to patients with myocardial or limb ischemia who have no other options. Although this is the group to target in the near future, it is not difficult to foresee a time when the sizeable population of patients who undergo bypass surgery but are not optimal candidates for that procedure, may be eligible for therapeutic angiogenesis, which might be performed at an earlier stage of disease and thus provide a greater possibility of a successful outcome.
Correspondence
Jeffrey M. Isner, M.D.St. Elizabeth’s Medical Center 736 Cambridge St.Boston, MA 02135 Tel: 617-789-2392 Fax: 617-779-6362 E-mail: jisner@opal.tufts.com