Therapeutic angiogenesis for management of refractory angina

INTRODUCTION — Advances in medical therapy and mechanical revascularization have significantly improved outcomes and the quality of life in patients with angina pectoris. In addition, the widespread application of drug-eluting stents has greatly expanded the ability of percutaneous coronary intervention (PCI) to treat patients with complex coronary anatomy. (See "Chronic coronary syndrome: Overview of care" and "Chronic coronary syndrome: Indications for revascularization" and "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention".)

Despite these advances, there are patients with angina that is refractory to medical therapy who are not candidates for traditional revascularization (PCI or coronary artery bypass graft surgery [CABG]) because of the inability to achieve complete revascularization or the high risk of CABG [1].

Such patients provide part of the rationale for novel therapeutic strategies to improve both prognosis and the quality of life. Therapeutic angiogenesis is one such option and will be reviewed here [2,3]. This approach is also being investigated for treatment of patients with peripheral artery disease, including both intermittent claudication and critical limb ischemia. (See "Investigational therapies for treating symptoms of lower extremity peripheral artery disease", section on 'Therapeutic angiogenesis'.)

Other investigational strategies, such as transmyocardial laser revascularization and new medications, are discussed separately. (See "Transmyocardial laser revascularization for management of refractory angina" and "New therapies for angina pectoris".)

GENERAL CONCEPTS — The goal of therapeutic angiogenesis is the induction of new coronary arterial vessels that can effectively provide blood supply to the area of myocardium subtended by diseased or occluded native coronary arteries. These "native bypass" vessels could then relieve myocardial ischemia, improve regional and global left ventricular performance, lessen symptoms of angina, and potentially improve patient prognosis [4].

Terminology — Before discussing the various strategies that have been evaluated for therapeutic angiogenesis, it is important to review the terminology that is used [3]. The term "angiogenesis" broadly refers to formation of new vessels. As currently understood, there are three major processes that result in new vessel formation:

●Vasculogenesis is a process of blood vessel formation from vascular progenitor cells. This is how the initial circulation forms during embryonic development. Studies in the 1990s and early 2000s suggested that there may be circulating endothelial progenitor cells (EPCs) that can form vasculature in mature adult tissue. However, more studies have demonstrated that EPCs are in fact either circulating adult endothelial cells or, most frequently, a subset of monocytes and that no vasculogenesis takes place in adult tissues [5]. These cells, however, may promote angiogenesis by the paracrine release of growth factors [6].

●True angiogenesis is de novo capillary growth stimulated by tissue injury, hypoxia, or ischemia.

●Arteriogenesis is a process of formation of new arteries from either pre-existing collaterals or by de novo arteriogenesis, a process that likely involves arterialization of the capillary bed [3].

The key drivers of angiogenesis are either ischemia, leading to a local activation of vascular endothelial growth factor (VEGF) expression due to activation of hypoxic signaling via a hypoxia-inducible factor (HIF) pathway, or inflammation that leads to local accumulation of VEGF via release from infiltrating inflammatory cells. Thus, the typical processes associated with angiogenesis are wound healing (tissue injury, inflammation) and ischemia (eg, ischemic limb). In both of these cases, angiogenesis serves a physiologic purpose by either supporting new tissue formation in wound repair or increasing blood supply in the case of ischemia.

Pathological angiogenesis occurs in the setting of excess local VEGF production that does not serve a physiologic function. Two prime examples are diabetic retinopathy, excess capillary proliferation in the retina due to inappropriate local release of VEGF, and tumor angiogenesis.

In contrast, the biology of arteriogenesis is much less well understood. It is not an ischemia-driven process, as collateral arteries form in the areas close to the interruption of a major arterial trunk (eg, an occluded common femoral or a proximal coronary artery) and not in the distal bed, which is actually ischemic. It seems clear that recruitment of certain circulating mononuclear cells is of critical importance. These cells probably serve as reservoirs of growth factors that promote arterial remodeling and/or de novo growth [7,8]. However, how they achieve this effect is not understood.

Restoration of arterial blood supply — Physiologic considerations strongly suggest that arteriogenesis, not angiogenesis, is the preferred type of neovascularization for restoration of myocardial perfusion [3,8].

The recruitment of certain circulating mononuclear cells is of critical importance. These cells probably serve as reservoirs of growth factors that promote arterial remodeling and/or de novo growth. The incorporation of such cells themselves into the growing vasculature is a low frequency event that is not likely to be of major physiological importance.

In a typical patient with atherosclerotic coronary disease, the arterial blood supply to a particular area of the myocardium is compromised because of a high grade stenosis or occlusion of the coronary artery supplying that region. The most effective strategy to restore the arterial blood supply would be to provide an alternative bypass around the occlusion. Since atherosclerotic lesions primarily affect epicardial coronary arteries, the required "bypass" should be in the same size range to be able to carry bulk flow.

It is theoretically possible to restore the blood supply by markedly expanding the capillary bed in the territory served by the compromised coronary artery. However, the degree of expansion required for this purpose is massive and may not be achievable [9]. Despite this limitation, no trial of therapeutic angiogenesis has specifically targeted arteriogenesis as its goal.

A number of angiogenic growth factors stimulate blood vessel growth, including fibroblast growth factors (FGFs), VEGFs, and platelet-derived growth factor (PDGF). However, none act independently and all have different and poorly understood biologic properties:

●VEGFs (in particular VEGF-A) are involved both in capillary growth induced by tissue ischemia and in arteriogenesis. VEGF expression is extremely sensitive to changes in local tissue oxygen levels, a process regulated by hypoxia-inducible factor (HIF)-1-alpha [10]. VEGF-A is usually present in limited quantities, thereby making regulation of protein levels extremely important for the biological function. Remarkably, even a 50 percent reduction in VEGF expression during development leads to lethality [11,12]. In addition to its role in angiogenesis, VEGF also plays a key role in arteriogenesis, although how it achieves that is poorly understood.

●FGFs (FGF1, FGF2, FGF4, and FGF5) are abundantly expressed by various cell types. Secretion and cell death leads to accumulation of these long-lived proteins in the extracellular matrix. Further, in contrast to VEGFs, their expression is not sensitive to ischemia or hypoxia. In contrast, the expression of FGF receptors, including both tyrosine kinase receptor FGFR1-R4 and a proteoglycan receptor syndecan-4 is regulated by ischemia [13]. FGFs are able to induce both capillary and arterial growth, thus participating both in angiogenesis and arteriogenesis. In contrast to VEGFs, FGFs appear to be particularly important in arteriogenesis with growing collaterals demonstrating abundant expression of FGF2 in surrounding monocytes [14].

CHALLENGES IN CLINICAL TRIALS — A few of the challenges facing researchers in developing effective revascularization strategies using biologic agents will be briefly summarized here.

Patient selection — The randomization of patients for therapeutic angiogenesis should be based upon age, sex, extent of hypercholesterolemia, and the extent of endogenous collateralization. This will require a sufficiently large number of patients (more than 400 per study arm) so that other unknown factors will be evenly distributed [15]. Patients with unstable angina, such as angina at rest, should be excluded. (See "Acute coronary syndrome: Terminology and classification".)

Cancer screening has been carried out in therapeutic angiogenesis trials. There is no compelling evidence that patients should undergo extensive testing for pre-existing malignancies solely because of a decision to start therapeutic angiogenesis. On the other hand, there are disease processes such as proliferative retinopathy that are known to be sensitive to vascular endothelial growth factor (VEGF) and proteinuric renal diseases such as membranous nephropathy that may be sensitive to fibroblast growth factors (FGFs). Such patients should be excluded in clinical trials evaluating systemic administration of these growth factors.

One important question is whether there are populations of patients that are unresponsive to angiogenic therapies, such as patients with diabetes who are resistant to VEGF signaling [16-18]. Unfortunately, there are currently no biomarkers available that would aid in selecting the correct population and in evaluating the therapeutic response.

Delivery of growth factors — How to effectively and safely deliver growth factors remains a major problem in therapeutic angiogenesis trials. Simple systemic (intravenous) infusions have either not been effective when used short term or have not been tried long term because of the fear of side effects. Methods of local delivery to the heart have included simple intracoronary infusion, which has been ineffective for both proteins and viruses; intramyocardial injections; pericardial administration; and locally applied polymer-based delivery.

It is not clear that any of these approaches is better than the others or that any are effective. Active research in this area includes retrograde coronary sinus delivery, and the development of effective delivery devices and noninvasive guidance techniques, such as electromechanical mapping that can distinguish between healthy and infarcted myocardium [19] and live three-dimensional echocardiography that accurately locates the tip of the delivery catheter in space [20].

Protein versus gene therapy — The emerging consensus is that effective angiogenic intervention will require the prolonged presence (perhaps four to six weeks) of the therapeutic agent at the desired site of action. This cannot be achieved with single-dose injections due to the short half-life of proteins in tissues. Repeat administration or the use of a sustained release polymer may prove effective.

Both plasmid and adenoviral based gene therapy approaches suffer from relatively short expression times (one to three days for plasmid, 7 to 14 days for adenovirus) and, with plasmid, relatively low expression levels.

Trial end points — Since the majority of patients currently being enrolled in therapeutic angiogenesis trials are selected on the basis of symptoms of angina and not life-threatening ischemia, quality-of-life measures should be the primary end point. Use of quality-of-life instruments such as the Seattle Angina Questionnaire may be appropriate [21].

Ideally, an angiogenesis-specific, quality-of-life assessment tool should be developed. Such instruments, which should be self-administered, must be reliable, validated, and scored by standard algorithms. Since there is a strong placebo effect with regard to quality of life, the patient, the referring physician, and the investigator must be blinded to the specific treatment [15].

Because of skepticism regarding subjective symptom assessment, much emphasis has been placed upon demonstration of physiologic improvements in myocardial perfusion and function. The goal is logical but has been technically challenging.

Exercise testing may or may not be a useful trial end point. Patients with variable baseline test results (>30 percent difference) as well as those with only minimally impaired exercise tolerance should be excluded from clinical trials. Pharmacological stress testing may be preferred to exercise protocols. Semiquantitative or quantitative single-photon emission computed tomography (SPECT) analysis is the most accepted tool at present.

Pro-angiogenic agents — The field has been dominated by the use of various vascular growth factors including VEGF, FGF and hepatocyte GF (HGF). As already discussed, a number of patients appear to demonstrate resistance to VEGF and perhaps other growth factors. Animal trials show that unlike responses in young, healthy animals, VEGF (and other factors) are much less effective in older atherosclerotic or diabetic animals [22]. If the VEGF-activated arteriogenic pathway was clearly understood, agents could be developed to activate it, bypassing VEGF receptor signaling resistance in this population.

VEGF

There are five members of the vascular endothelial growth factor (VEGF) family, with VEGF-A being the most extensively studied. VEGF therapy has been evaluated in studies administering the VEGF protein and in gene therapy trials in which the gene for VEGF was administered in a plasmid or viral vector.

VIVA trial — Although the initial open label phase I of VEGF-A intracoronary infusion in patients with coronary heart disease (CHD) was interpreted as positive, this was not confirmed in VIVA, a randomized, double-blind, placebo-controlled phase II trial of intracoronary recombinant human VEGF (VIVA) [23]. In this trial, 178 patients were assigned to placebo or to VEGF given at either low or high doses. The infusions were given into coronary arteries for 10 minutes, followed by three repeat intravenous infusions over the following nine days. The following findings were noted:

●At 60 days, there were no significant differences in exercise tolerance or angina class between the treated and control groups.

●At 120 days, patients receiving high dose VEGF had a significant improvement in angina class compared to the other two groups. However, nuclear perfusion imaging showed no differences among the three groups, raising doubt about the improvement in angina. This suggests that angina class may not be a good end point for trials of this intervention.

KAT trial — The KAT trial randomly assigned 103 patients with symptomatic myocardial ischemia to VEGF DNA in a plasmid liposome, an adenoviral vector, or control via intracoronary infusion at the time of percutaneous coronary intervention (PCI; 90 percent with stents) [24]. At six months, a significant improvement in myocardial perfusion using stress technetium-99m sestamibi SPECT imaging was seen in patients receiving adenoviral VEGF, but not those treated with plasmid VEGF or controls. There were no significant improvements in exercise time, functional class, or restenosis rate. However, the use of PCI and the small numbers of patients make it difficult to interpret these findings.

Euroinject One trial — The Euroinject One trial randomly assigned 80 patients with advanced CHD but not further medical or revascularization options to receive either phVEGF-A (165) or placebo plasmid in the myocardial region showing stress-induced myocardial perfusion defects [25]. At three months, there were no significant differences between the groups on analysis of myocardial stress perfusion defects or on semiquantitative analysis of the change in perfusion in the treated region.

NORTHERN Trial — The trial enrolled 93 patients with refractory class 3 or 4 angina and randomly assigned them to injection VEGF-A plasmid or placebo. There was no difference in VEGF-treated or placebo-treated patients after three or six months with regard to SPECT-assessed ischemic burden. Interestingly, both groups demonstrated a significant reduction in the ischemic burden over time.

Other members of VEGF family, such as VEGF-B and PlGF, have been tested in pre-clinical setting, but to date, there are no clinical studies of their efficacy.

FIBROBLAST GROWTH FACTORS

There are 22 members of the fibroblast growth factors (FGFs) growth factor family. FGF-1, FGF-2, and FGF-4 have received the most extensive evaluation to date.

FGF-2 — FGF-2 (basic FGF or bFGF) has been evaluated in several clinical trials. The largest trial, the FGF Initiating Revascularization Support Trial (FIRST) trial, randomly assigned 337 patients who were ineligible for angioplasty or revascularization to placebo or one of three doses of intracoronary recombinant FGF-2 (0.3, 3, or 30 µg/kg) [26]. The following findings were noted:

●At 90 days, there was no significant improvement in exercise treadmill time or nuclear perfusion parameters among the treatment groups. However, there was an improvement in patient and physician perception of angina as assessed by the Seattle Angina Questionnaire and the Canadian Cardiovascular Society angina classification (table 1).

●At 180 days, there were no significant differences between the treatment groups.

A smaller trial of 24 patients evaluated the efficacy of local application of an FGF-2 containing polymer at the time of bypass surgery to the area of myocardium that could not be revascularized during the procedure. Patients receiving high-dose of FGF-2 (100 mg) had a significant improvement in perfusion of the target area compared to low-dose FGF-2 (10 mg) or placebo controls [27]. Three years later, the patients receiving high-dose FGF-2 had significantly less angina than placebo controls [28]. Although encouraging, interpretation of these results is clouded by the presence of concomitant bypass surgery.

FGF-1 — The safety and efficacy of intramyocardial FGF-1 (acidic FGF or aFGF) was evaluated in 20 patients with three vessel disease undergoing coronary artery bypass graft surgery (CABG); FGF-1 was injected at the site of the left internal mammary artery to left anterior descending artery anastomosis [29]. Coronary angiography performed two weeks after FGF-1 injection showed a capillary blush at the site of injection in a majority of patients. No significant side effects were noted.

FGF-4 — A series of AGENT trials examined the safety and efficacy of intracoronary injections of an adenoviral-encoded FGF-4 gene. The AGENT-1 trial enrolled 79 patients who were randomly assigned to various doses of Ad-FGF4 [30]. Patients in two dose groups had a trend toward a greater increase in exercise time at four weeks compared to the placebo group (1.3 versus 0.7 minutes), while patients with a baseline exercise time ≤10 minutes had a significant increase in exercise time (1.6 versus 0.6 minutes).

The subsequent AGENT-2 trial of FGF-4 gene transfer evaluated 52 patients with stable angina and reversible ischemia on adenosine technetium-99m sestamibi SPECT imaging [31]. Patients who received gene therapy had a nonsignificant reduction in the size of the ischemic defect compared to no change among those receiving placebo.

Two large double-blind phase III trials of Ad-FGF4 (AGENT-3 and AGENT-4) were negative for their primary end point, although an over-optimistic analysis suggested a benefit in a certain population of middle-aged women [32]. There are, however, no biological reasons to support this conjecture.

A particular issue that probably doomed the Ad-FGF4 strategy was its intracoronary administration. Animal studies show very little myocardial uptake of an intracoronary-infused adenovirus in intact arterial beds [33]. Thus, this strategy was doomed to failure because of inefficient delivery. On the other hand, the AGENT program is a clear demonstration of the safety of short-term adenoviral therapy in a large cohort of patients.

GM-CSF — Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a cytokine secreted by a wide variety of cell types that acts broadly upon hematopoietic cells. Activated macrophages can induce vascular proliferation, suggesting a role for the administration of recombinant human GM-CSF (rhGM-CSF). (See "Introduction to recombinant hematopoietic growth factors".)

The possible efficacy was examined in an initial trial of 21 patients with severe inoperable coronary disease who were randomly assigned to placebo or intracoronary rhGM-CSF followed by two weeks of subcutaneous drug [34]. Compared to placebo, rhGM-CSF improved collateral flow, measured by intracoronary sensor guidewires, and reduced electrocardiogram signs of ischemia during coronary balloon inflation.

HEPARIN AND ADENOSINE — In animals, heparin can accelerate the formation of coronary collaterals induced by ischemia [35]. Clinical studies have shown that the use of heparin to treat repeated episodes of exercise-induced ischemia can improve myocardial perfusion and reduced electrocardiographic evidence of ischemia [36].

Various models of ischemic preconditioning, which refers to the protection conferred to ischemic myocardium by preceding brief periods of sublethal ischemia, have implicated the involvement of adenosine, which is released locally during ischemia and may have a protective role. (See "Myocardial ischemic conditioning: Pathogenesis".)

The role of heparin and adenosine as treatment for refractory angina was evaluated in a study of 21 patients with chronic stable angina that was refractory to conventional medical therapy; the patients were not candidates for revascularization [37]. Heparin (10,000 units) and adenosine (140 µg/kg per min for six minutes) were administered daily for 10 days. Compared to placebo, heparin and adenosine reduced the extent (by 9 percent) and severity (by 14 percent) of myocardial perfusion abnormalities on imaging.

POTENTIAL COMPLICATIONS — There are a number of potential complications associated with therapeutic angiogenesis:

●Aberrant vascular proliferation in adjacent and perhaps distant nontargeted tissues.

●Increased vascular permeability; the ensuing extravasation of plasma and plasma proteins into tissues can trigger the clotting system, leading to the deposition of fibrin gel and the development of edema, induction of new blood vessels, and fibroblast activation.

●Triggering of growth of coexisting, but unrecognized, neoplasms or the development of de novo tumors.

●Proatherogenic effects, including an increase in neointimal smooth muscle cell proliferation, neointimal mass, and vaso vasorum supplying the atherosclerotic lesions. As an example, vascular endothelial growth factor (VEGF) promotes monocyte activation and migration and can induce a local inflammatory reaction, changes that could promote atherosclerosis and plaque development. Consistent with this hypothesis are the observations in atherosclerosis-prone animals that human recombinant VEGF increased the rate and degree of atherosclerotic plaque formation in the thoracic aorta and that inhibitors of angiogenesis reduced intimal neovascularization and plaque growth [38].

●Hazards associated with viral vectors, as the transfection of cells with replication-incompetent adenoviral vectors may induce immune or inflammatory responses.

●Hazards associated with direct myocardial delivery of angiogenic factors.

Despite these potential complications, few significant complications have been observed in angiogenesis trials to date.

SUMMARY — There are patients with angina that is refractory to medical therapy who are not candidates for traditional revascularization (percutaneous coronary intervention [PCI] or coronary artery bypass graft surgery [CABG]) because of the inability to achieve complete revascularization or the high risk of CABG. Therapeutic angiogenesis is under investigation as one option for improving the quality of life in these patients.

The goal of therapeutic angiogenesis is the induction of new coronary arterial vessels that can effectively provide blood supply to the area of myocardium subtended by diseased or occluded native coronary arteries. A number of angiogenic growth factors stimulate blood vessel growth including fibroblast growth factors (FGFs), vascular endothelial growth factors (VEGFs), and platelet-derived growth factor (PDGF) and many of these are under investigation. To date, none of them have been shown to have clinical utility.

The difficulty in translating experimental successes with pro-angiogenic therapy to clinical practice probably stems from three key sources: ineffective delivery (especially in myocardial angiogenesis trials), the presence of co-morbid conditions (diabetes, atherosclerosis, etc) that can inhibit angiogenic signaling [18], and advanced age of patients in trials [39]. Finally, it is also quite likely that choosing the "no-option" population for these trials results in selection of patients with genetic defects preventing new vessel formation.