Critical determinants of limb ischemia

Clinical manifestations of PAD

Peripheral arterial disease (PAD) involving the lower extremities affects 8–12 million Americans(1). Typically, PAD is secondary to atherosclerotic occlusive disease of the aorto-iliac, femoral, popliteal and/or the infrapopliteal arteries. The clinical manifestations depend in part upon the level of disease and its severity. The classic symptom of intermittent claudication (pain with walking, relieved by rest) typically involves the calf, but the thigh and/or buttocks can be affected with more proximal arterial disease. Classic symptoms are only manifested by 10–30% of patients (2,3). More often, the patients are asymptomatic, or their symptoms are disguised by (or attributed to) associated diabetic neuropathy or musculoskeletal disorders, each of which are common in the demographic affected by PAD.

A more severe form of PAD is characterized by pain at rest (typically in the foot, occurring often at night, and relieved by dependency). These individuals have a limb in jeopardy, and the most minor trauma (from a poorly fitting shoe, or a carelessly clipped toenail) can result in a non-healing wound, infection and amputation. This form of PAD is known as critical limb ischemia, and affects about 10% of PAD patients. Critical limb ischemia is a mortal illness: at one year following presentation, 30% have suffered an amputation and 25% of patients are dead (4).

Why are there differences in the clinical presentation of PAD?

Although hemodynamically significant obstructions of the conduit arteries play an acknowledged role in the manifestations of PAD, there is considerable heterogeneity in the degree of symptoms. Furthermore, there is only a modest correlation of hemodynamic assessments, such as the ankle-brachial index (ABI; the pressure at the ankle divided by the pressure at the arm), with functional capacity as assessed by treadmill testing (5). This heterogeneity may be due in part to individual differences in the response of the skeletal muscle to ischemia. For example, the expression and activity of key mitochondrial enzymes are altered in skeletal muscle from PAD patients, and mitochondrial electron transport is impaired (6). Metabolic byproducts of fatty acid oxidation, the acylcarnitines, accumulate in both the plasma and skeletal muscle of patients with PAD, and are strongly correlated with impairment in exercise capacity (6).

The vascular response to conduit vessel occlusion also appears to be subject to individual variation. Measures of skin microcirculation have been used prospectively to predict amputation in CLI patients with unreconstructable vascular disease (7). In patients with reduced skin capillary density (< 20/mm²), low transcutaneous oxygen (TcPO2 < 10 mmHg), and absent reactive hyperemia, limb survival at one year was only 15%, compared to 88% in CLI patients with greater values for the skin microcirculation.

Furthermore, it is widely recognized that collateral vessel formation is heterogeneous in patients with the same degree of conduit vessel obstruction. In one patient with a superficial femoral artery occlusion, robust collateral formation (arteriogenesis) via the deep femoral artery and geniculate collaterals will revascularize the infrapopliteal vessels. The patient may have minimal or no symptoms. Another patient will be significantly limited by the same occlusive disease, in a setting of sparse collateral formation.

Regulation of angiogenesis

What is responsible for the heterogeneity of vascular regeneration in PAD patients? An important clue has been provided by the work of Findley and colleagues (8). These investigators studied healthy controls and patients with peripheral arterial disease (PAD) manifested as intermittent claudication (IC) or as critical limb ischemia (CLI). In these patients they measured plasma levels of proteins from two families of angiogenic growth factors, the vascular endothelial growth factors (VEGF) and the angiopoietins (Ang).

The vascular endothelial growth factors (VEGF-A, VEGF-B, VEGF-C, VEGF-D and Placental growth factor) and the angiopoietins (Ang 1 and 2) are major mediators of angiogenesis and lymphangiogenesis (9). The receptors for the VEGFs include VEGFR-1, VEGFR-2 and VEGFR-3, whereas the angiopoietin receptors include Tie 1 and Tie 2. The diversity of angiogenic factors and their receptors permits fine-tuning of the angiogenic response as new capillaries form and differentiate, and allows for specificity.

VEGF-A is the prototype member of the VEGF family. The loss of a single allele of VEGF-A is embryonically lethal, due to failure of hematopoiesis and vasculogenesis. VEGF-A has four isoforms, including the one analyzed in the Findley study, VEGF₁₆₅. The various isoforms of VEGF induce endothelial cell proliferation, migration, and capillary tube formation. The angiopoietins are also necessary for angiogenesis, modulating endothelial cell migration, adhesion and survival, acting through their endothelial receptor Tie2. Genetic disruption of Tie 2 also results in embryonic lethality, due to vascular defects(10).

Findley and colleagues noted higher plasma levels of the angiogenic cytokines Ang 2 and VEGF in the PAD patients. Furthermore, plasma levels of VEGF were significantly greater in those with more severe disease, ie. in patients with CLI. The levels of Ang 2 also tended to be higher in the CLI patients (the lack of statistical significance was likely a type II error due to insufficient numbers of patients in this small study). These findings are consistent with the notion that greater ischemia is associated with greater activation of the angiogenic response to hypoxia. Parenthetically, much of the hypoxic response is under the control of the transcriptional factor, hypoxia-inducible factor 1α (HIF-1α). At low oxygen tension, HIF-1α is stabilized and can translocate to the nucleus to orchestrate the genetic response to ischemia, activating genes encoding angiogenic cytokines, metabolism, vasodilation and cell survival (11).

Endogenous inhibitors of angiogenesis

Of relevance to this study, it is known that the angiogenic receptors VEGFR-1 and Tie2 can be released by endothelial cells into the circulation. These solubilized receptors can still bind their respective growth factors. By scavenging VEGF or Ang, these soluble receptors may act as a brake on angiogenesis. The soluble receptors may represent a form of negative feedback, as VEGF is known to promote the endothelial shedding of Tie 2 (12). Notably, overexpression of either of these soluble receptors in animal models inhibits tumor angiogenesis and tumor growth (13, 14). The logical extension of these fundamental insights, a soluble VEGF receptor for inhibition of tumor angiogenesis and growth, is showing promise in early clinical trials (15).

Could endogenous soluble receptors, scavenging angiogenic cytokines, play an adverse role in the angiogenic response to PAD? Intriguingly, Findley and colleagues also observed an increase in the plasma levels of the soluble Ang receptor (sTie2) in the PAD patients. Furthermore, the plasma levels of sTie2 were highest in the most symptomatic patients. The greater elevation of sTie2 in the CLI patients remained after controlling for differences in CV risk factors or ABI. These findings suggest that increased levels of sTie2 correlate with the severity of disease, independently of the hemodynamic impairment.

One might speculate that the elevation of the soluble receptor impairs angiogenic response in these patients. To confirm this hypothesis, it would be useful to study a larger group of PAD patients, including asymptomatic to most symptomatic subjects. One would correlate plasma levels of the angiogenic cytokines and their soluble receptors to an angiographic assessment of collaterals, as well as capillary density in needle biopsies of calf muscle. Individuals with greater levels of the angiogenic cytokines, and lower levels of the soluble receptors, would be expected to have greater angiogenic response.

New insights into angiogenic modulation

The notion that individual patients may have different responses to the same ischemic stimulus finds support in the coronary circulation. Recently, a genetic variant of HIF-1α has been described. Patients with coronary artery disease, that also had this genetic variant, were observed to have fewer coronary collateral vessels (16). Undoubtedly, other genetic factors will be found to play a role in the heterogeneity of vascular regeneration in response to ischemia. Indeed, the Duke group is continuing their hunt in C57Bl/6 mice, where a segment of chromosome 7 containing 37 genes confers resistance to limb loss after femoral artery ligation (17). One or more of these genes may be involved in the enhanced angiogenic response and superior perfusion observed in this strain of mice. Taken together, the translational studies of Findley and colleagues will provide new clues to the mechanisms underlying impaired vascular regeneration in PAD, and may lead us to new therapeutic approaches to promote arteriogenesis and to enhance angiogenesis.