INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE ANGIOGENIC GROWTH FACTORS are a group of cytokines, the actions of which in vascular development/remodeling could provide therapeutic benefit in peripheral arterial insufficiency. They include the heparin-binding proteins, acidic and basic fibroblast growth factor (aFGF and bFGF), and vascular endothelial growth factor (VEGF), the administration of which leads to vascular remodeling in vivo and expected or demonstrated improvements in the hemodynamic deficit introduced by the experimental vascular occlusion. Enlargement and/or expansion of large upstream vessels (2-4, 22, 26) associated with endothelial cell proliferation (9, 21) should reduce collateral resistance and improve blood flow downstream to the tissue at risk of ischemia. Blood flow to muscle, affected by experimental vascular occlusion, is significantly improved with VEGF (24) and bFGF (26) administration. This leads to an improvement in muscle performance during contractions in situ (26).

Previous work evaluating the angiogenic growth factors has appropriately focused on establishing and characterizing the nature of the in vivo response, without systematic assessment of growth factor efficacy. The most common strategy for administration has been in or near the ischemic locus with single or multiple bolus deliveries of the growth factor, sometimes in increasing amounts. However, recent evidence shows that the magnitude of the increase in collateral flow, induced by bFGF, is dependent on the duration of in vivo infusion over the first 2 wk of treatment, but not thereafter (26). Interestingly, delivery close to the ischemic region is not essential, inasmuch as systemic delivery of growth factors appears to be efficacious (3, 11, 24). The absence of more knowledge is unfortunate, since an understanding of the dosing routes and regimens provides insight into therapeutic use of the growth factors. Furthermore, the flow benefit realized with bFGF-induced vascular remodeling in experimental peripheral arterial insufficiency has not been assessed in vivo during physiologically relevant activity, such as treadmill exercise. The present study was designed to evaluate the dose response for bFGF, the importance of critical routes and dosing regimens, and the specificity of action in the vasculature affected by occlusion.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental design. Adult male Sprague-Dawley rats (325-350 g; Taconic Farms, Germantown, NY) were subjected to bilateral femoral artery ligation and randomly assigned to a treatment group described below. Sixteen days later collateral-dependent limb blood flow was measured at two speeds during treadmill running. Table 1 summarizes the three experimental series that evaluated different dosing regimens, each in a blinded manner. In addition, a dose response for intra-arterial infusion into the left iliac artery was completed by using a high dose (50 µg · kg¹ · day¹ for 14 days) and lower doses (0.5 and 0.05 µg · kg¹ · day¹ for 14 days) than used in the primary series.

1

1

1

1

Animal care. Rats were housed three per cage in a temperature-controlled room (20°C) with a 12:12-h light-dark cycle. Animals were given Purina rat chow and tap water ad libitum. All procedures and the care and treatment of animals were approved by the Committee for the Humane Use of Animals of the State University of New York Health Science Center ( Syracuse, NY).

Surgical procedures. Bilateral ligation of the femoral artery was performed under anesthesia with ketamine (100 mg/kg) and acepromazine (0.5 mg/kg), as done previously (26). A ligature was securely tightened around each femoral artery isolated just distal to the inguinal ligament. Catheters connected to an osmotic pump (Alzet) were inserted into the iliac artery, for infusion near the site of collateral expansion, or into a branch of the right jugular vein, for systemic delivery of vehicle or bFGF. Topical antibiotic powder (Neo-Predef, Upjohn) was placed on each wound before closure with skin clips. This surgery produces a uniform peripheral arterial insufficiency, and the animals quickly recover with 100% success.

Blood flow measurement. Surgical preparation for blood flow measurement was performed early in the day, with blood flow determinations made later (4-6 h) in the afternoon, as done previously (6, 15, 31). A catheter was placed in the caudal artery to obtain the reference blood sample during microsphere infusion. A second catheter was inserted into the left carotid artery and extended to the arch of the aorta for monitoring blood pressure and heart rate and for infusion of the microspheres. Both catheters were filled with heparin-saline (100 IU/ml), led under the skin, and exteriorized on the back at the base of the neck.

1

1

Materials. Recombinant human bFGF (Scios) or vehicle [PBS containing 1.6% glycerol to stabilize protein, 0.02% sodium azide as a bacteriostat, and 10% saturated sodium citrate for anticoagulation, as described by Liu et al. (13)] was delivered with a miniosmotic pump (Alzet) or by injection (intravenous or subcutaneous), as appropriate for the experimental group. Infusion rates were 0.5-1.0 µl/h, depending on the specific pump chosen to achieve the desired duration of treatment.

Analysis of the data. Statistical analyses included repeated-measures ANOVA with individual mean comparisons by Tukey's procedure (20). Significant differences for main treatment effects and individual comparisons were recognized at P < 0.05. Values are means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Blood flow determinations were generally achieved without complications. In a few cases (~10%) the second blood flow determination could not be obtained. These are identified in Tables 3-5.

Heart rates and blood pressures were elevated during exercise, compared with preexercise, but were not different at the two running speeds (Table 2). Renal blood flows decreased when the treadmill speed was increased (Table 3). Renal blood flows shown in Table 3 are a combination of flows across kidneys for each determination, since flows were well matched across kidneys (r = 0.898). Blood flows to the right kidney averaged 98.9 ± 0.96% (n = 296) of blood flows to the left kidney. This indicates good mixing of the microspheres during all the blood flow determinations in these experiments. Blood flows to the hindlimb muscles were similar at the two running speeds for each group (Table 4). Thus the data have been combined for presentation.

Results for animals treated similarly in the three series of experiments were not different. Therefore, the results have been combined as noted in Table 2; this accounts for the unequal group sizes identified in Tables 2-5.

Continuous intra-arterial treatments. Animals receiving intra-arterial bFGF treatment (5 µg · kg¹ · day¹) for 3 or 14 days exhibited greater blood flow to the total hindlimb than vehicle-treated controls measured 16 days after bilateral femoral artery ligation (P < 0.001; Table 4). Increased flows to nearly all muscle groups in proximal and distal regions were observed (Table 5). The largest benefit was measured in the calf muscles (Table 4). These increases in collateral-dependent blood flows were observed with similar aortic pressures across groups (Table 2). There was no statistically significant difference between blood flow in animals treated with bFGF for 3 days and blood flow in animals treated for 14 days.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Correlations among blood flows to corresponding individual muscle sections (see Table 5) across limbs of vehicle (A)- and basic fibroblast growth factor (bFGF, B)-treated groups. Some SE bars are hidden by symbols or not shown for clarity.

1

1

1

1

1

1

1

1

1

1

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Dose response of collateral blood flow increase induced by bFGF infusion into iliac artery for 14 days. Similar increases in collateral blood flow (~80 ml · min1 · 100 g1) were observed with intravenous injections of bFGF (Table 4). Blood flows were determined during treadmill running on day 16. Blood flows for vehicle and 5 µg · kg1 · day1 groups are composite values obtained in 3 dosing regimen experiments. Numbers in parentheses represent number of animals.

Continuous intravenous treatment. Animals receiving intravenous bFGF (5 µg · kg¹ · day¹) for 3 days exhibited greater blood flow in the proximal and distal hindlimb regions (P < 0.001; Table 4), as well as individual muscle groups within those regions (Table 5), with the greatest effect in the calf muscles. There were no statistically significant differences among the groups receiving bFGF.

Bolus intravenous and subcutaneous treatments. Animals receiving intravenous bolus bFGF treatment (15 µg/kg) on days 0 and 7 exhibited an increase in blood flow to proximal and distal muscle groups compared with vehicle treatment (Table 4). The stimulatory effect of bFGF on collateral-dependent blood flow was similar among all groups, whether bFGF was administered by infusion (intra-arterial or intravenous) or by bolus intravenous injection on days 0 and 7 (Tables 4 and 5).

Specificity of bFGF actions. Collateral-dependent blood flow to the calf muscles is least after acute, same-day occlusion of the femoral arteries (Fig. 3). There is a modest increase in calf muscle blood flow with time of administration of vehicle (Fig. 3). When bFGF is administered, there is a marked increase in collateral-dependent blood flow; however, it requires coincident occlusion of the femoral artery, inasmuch as blood flow in the limb that was not occluded during the period of bFGF infusion did not increase above that observed for acute same-day occlusion.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Specificity of bFGF effect. Collateral blood flows during treadmill exercise to calf muscles of animals 16 days after bilateral occlusion of femoral arteries and infusion of bFGF (5 µg · kg1 · day1); 16 days after occlusion of left femoral artery, acute occlusion of right femoral artery, and infusion of bFGF (5 µg · kg1 · day1); 16 days after occlusion of left femoral artery, acute occlusion of right femoral artery, and infusion of vehicle; and after acute same-day occlusion of both femoral arteries are shown. Note absence of bFGF-induced increase in calf muscle blood flow without concurrent occlusion of femoral artery. Numbers in parentheses represent number of animals. * Significantly different from all other blood flows (P < 0.05).

Separation of the time of vascular occlusion and the initiation of bFGF delivery. There was a robust increase in collateral blood flow to the calf muscles, even though bFGF administration was delayed by 2 wk after occlusion of the femoral artery (Fig. 4).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Collateral blood flows during treadmill exercise to calf muscles of animals that began receiving bFGF at time of vascular occlusion (solid circle) or after 2-wk delay (gray circle) compared with appropriate vehicle control groups (open circles). Calf muscle blood flow measured on same day as femoral artery occlusion (acute; time 0) is illustrated for comparison. Numbers in parentheses represent number of animals. * Significantly different from time-matched vehicle control groups (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first report demonstrating the effectiveness of intravenous bFGF administration in a model of experimental vascular insufficiency. Previous studies have demonstrated beneficial effects of perivascular (5), intramuscular (2, 22), or intra-arterial (4) bFGF and intravenous (24) or intra-arterial (3, 21, 22) VEGF delivery protocols with experimental vascular occlusion. Of particular significance in the present study is the unique means of evaluating collateral-dependent blood flow in conscious, active animals during treadmill running. The distal tissue of the hindlimb is most at risk of ischemia with occlusion of the proximal femoral artery (9, 27). Flow capacity is greatly reduced, by ~80-90% (27), but remains two- to threefold above the flow needs of resting quiescent muscle (14). Thus tissue ischemia is induced only when flow demands increase, as, for example, during exercise. This establishes an intolerance to exercise (29, 30). In this regard, the present model of experimental arterial insufficiency attempts to simulate acute-onset intermittent claudication and avoids the more severe conditions of ischemia at rest and/or tissue necrosis that can occur with more extensive arterial obstruction (5, 18). We measured muscle blood flow twice, during a challenging running speed and, after a brief rest, during a slightly higher treadmill speed that is expected to further dilate the vasculature of the active muscle (e.g., calf muscles) (10). The absence of any further increase in blood flow to the calf muscles during the second, higher running speed indicates that the upstream resistance of the collateral vessels was limiting and thereby determined downstream blood flow. We previously characterized enlargement of these upstream conduit vessels by X-ray images of arterial casts (26). Thus we interpret these calf muscle blood flows as representing improvements in maximal collateral-dependent blood flows.

The increases in collateral blood flow with bFGF are physiologically relevant. Although we have not evaluated their impact on exercise tolerance in this study, we previously observed increases in calf muscle performance that were proportional to the increase in measured muscle blood flow, in animals with enhanced collateral blood flow induced by bFGF administration (26), by increased physical activity (30), and by a combination of bFGF administration and enhanced physical activity (29). Thus an improvement in exercise tolerance during a standard exercise test in vivo would be expected in the present bFGF-treated animals. Similarly, the greater renal blood flows observed in all the bFGF-treated animals could reflect a lower sympathetic outflow during exercise. Although not evident in blood pressures, this may result from a lesser ischemic pressor response associated with the better-perfused active ischemic muscle (28).

Using this model, we showed that intra-arterial administration of bFGF for 2 wk at 5 or 50 µg · kg¹ · day¹ was equally effective in stimulating collateral-dependent blood flow. In contrast, 2 wk of continuous intra-arterial bFGF at 0.5 µg · kg¹ · day¹ was less effective than 5 µg · kg¹ · day¹. Thus maximally effective and no-effect doses were defined. In addition, we show that dosing regimens that are short in duration (e.g., bolus) were as effective as protocols involving considerably longer durations of treatment (e.g., 3 days or 2 wk). Intravenous and intra-arterial dosing protocols were equally effective. The only exception was with the subcutaneous protocol, which provided a less dramatic, albeit significant, increase in collateral blood flow, possibly related to the dose administered and/or the pharmacokinetics of the subcutaneous route. Thus systemic administration of bFGF was fully effective at increasing collateral blood flow.

The improvements in collateral blood flows were independent of the duration of bFGF administration; for example, the 1-, 3- and 14-day intra-arterial protocols yielded similar increases in calf muscle blood flow. A number of factors could contribute to this outcome. First, the measured increases in collateral blood flow could represent a saturation of the vascular change in this model. The further expansion of the bFGF-induced increase in collateral blood flow by routine physical activity (29), however, argues that other factors are likely involved. Second, initial events associated with the onset of bFGF treatment could be determining factors. For example, in all our experiments, the onset of bFGF treatment was always coincident with the introduction of vascular occlusion by ligation of the femoral artery. It is known that, shortly after vascular occlusion, nearby vessels that likely serve as subsequent collateral conduits exhibit proliferative activity necessary for vascular remodeling (9), even in the absence of angiogenic growth factor therapy. We provide evidence, however, that the improvement in collateral blood flow with bFGF administration is robust, even when the initiation of bFGF delivery is separated from vascular occlusion by a substantial time (Fig. 4). Third, bFGF possesses heparin-binding characteristics that could influence special distribution-activity relationships to impact its biological effectiveness (16, 17, 23). The present findings are different from our previous work, in which bFGF induced time-dependent increases in collateral blood flow over the first 2 wk of intra-arterial infusion, but not thereafter (26). In that study, bFGF was complexed with heparin before administration. This would extend its circulation time (25) and possibly alter its clearance and deposition by the heparan sulfates of the extracellular matrix. In the present study, bFGF was administered in the absence of exogenous heparin, thereby making it amenable to rapid clearance to the extracellular matrix for "storage" and possible subsequent action in vascular remodeling. Knowledge of the definitive basis for the time independence awaits further study.

Previous studies in animals with femoral artery occlusion have shown that bFGF treatment results in an expansion of upper thigh vessels, likely serving as collaterals, as judged by X-ray films of arterial casts (9, 26). Ito et al. (9) studied the vascular response to femoral artery ligation in rabbits in the absence of any treatment. Within 3 days of femoral artery occlusion, well-defined collateral vessels were observed in the proximal muscle region (thigh) and not in the distal muscle region of the hindlimb. However, significant proliferation of endothelial cells and smooth muscle cells was also reported in the thigh, indicating that collateral lumen enlargement was not simply vasodilatation but involved neovascular development, a process termed arteriogenesis by Schaper and Ito (9, 19). In the lower limb, where the greatest perfusion deficits were described, endothelial cell proliferation in the absence of smooth muscle cell proliferation was reported, suggesting that angiogenesis predominates over collateral remodeling in this region. This probably accounts for the increase in muscle capillary density that can be found in bFGF-treated animals (1, 26). Inasmuch as bFGF is a potent mitogen for endothelial and smooth muscle cells, it is possible that bFGF potentiates vascular remodeling of existing collaterals in the proximal region, the development of new collaterals between the proximal and distal regions, and angiogenesis in the distal hindlimb region.

Another interesting response revealed by our data set is the similar increase in collateral blood flow across the hindlimbs of each animal. Although blood flows to specific muscle sections within a single hindlimb can be different, because of different vascular conductances of the "red" and "white" muscle regions, each corresponding tissue of both hindlimbs exhibited nearly identical increases in blood flow with each bFGF treatment (Fig. 1). This was not the case in the absence of femoral artery occlusion in one leg (Fig. 3). When an effective dose of bFGF was administered to animals that experienced only unilateral femoral artery occlusion, blood flow to the calf muscles of the contralateral limb, the femoral artery of which was not occluded throughout the 14 days of bFGF administration, was not significantly elevated. Similar to the work of Lindner et al. (12), the normal vessels remain unresponsive to bFGF. To our knowledge, this is the first evidence showing that existing vessels that ultimately serve as collateral conduits are unresponsive to circulating bFGF, unless vascular obstruction is present. Thus vascular remodeling was specific to the region affected by occlusion of the femoral artery. This implies that site-specific alterations occurred in the tissue to become responsive to bFGF. Although this process of arteriogenesis is not well understood (8), it is not likely related to metabolic consequences of ischemia, inasmuch as the tissue surrounding the collateral conduits receive a relatively high blood flow (proximal in Table 4) (5, 9, 26), whereas the tissue at risk of ischemia is the far distant muscle of the lower limb. Rather, hemodynamic changes related to increased flow velocity, wall stress and shear stress (8), and/or monocyte involvement (1) in the remodeling vessels are likely important. A simple testable hypothesis is that growth factor receptors in the endothelium may be upregulated by the altered flow dynamics caused by occlusion of the femoral artery. Recognition that responsiveness of existing vessels to bFGF appears only in the region affected by vascular occlusion has practical utility and raises the prospects for therapeutic use of bFGF in patients with peripheral arterial insufficiency that is specific to the region of need.