Comparative Effects of Basic Fibroblast Growth Factor and Vascular Endothelial Growth Factor on Coronary Collateral Development and the Arterial Response to Injury
Background We have shown that the angiogenic peptides basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) enhance canine coronary collateral development when administered for ≥4 weeks. bFGF, a pluripotent mitogen of mesodermally derived cells, could theoretically exacerbate neointimal smooth muscle cell hyperplasia, a fundamental component of atherosclerosis. VEGF, an endothelial cell–specific mitogen and vascular permeability factor, could have deleterious effects related to vascular hyperpermeability. The present investigation had two aims: (1) to ascertain whether brief (7-day) systemic arterial treatment with bFGF or VEGF would improve myocardial collateral perfusion and (2) to determine whether these peptides induce neointimal accumulation in vivo.
Methods and Results Dogs were subjected to ameroid-induced occlusion of the left circumflex coronary artery and randomized to bFGF 1.74 mg (n=9), VEGF 0.72 mg (n=9), or saline (n=10) as a daily left atrial bolus (days 10 to 16). Additional dogs were randomized to VEGF 0.72 mg (n=6) or saline (n=5); however, treatment was delayed by 1 week. Coincident with the institution of treatment, all dogs underwent balloon denudation injury of the iliofemoral artery. bFGF markedly increased maximal collateral flow but did not exacerbate neointimal accumulation. VEGF had no discernible effect on maximal collateral flow, but it exacerbated neointimal thickening after vascular injury.
Conclusions Short-term treatment with bFGF enhanced collateral development without increasing neointimal accumulation at sites of vascular injury. Although VEGF did not increase collateral development as administered in this study, it significantly exacerbated neointimal accumulation. These data provide support for the clinical investigation of bFGF in selected patients with ischemic heart disease.
Vascular endothelial growth factor and bFGF are angiogenic heparin-binding peptides that exert a number of trophic effects on vascular target cells.1 2 These peptides may be critically important in vasculogenesis, atherogenesis, and vascular remodeling in response to injury. VEGF is a 46-kD dimeric peptide with target cell specificity for endothelial cells in vitro.1 3 4 The peptide was first isolated from bovine pituitary follicular cells5 and stimulates endothelial cells to proliferate and to express collagenase in vitro.6 VEGF is a secreted peptide with a hydrophobic signal sequence7 and is synonymous with vascular permeability factor; therefore, it has the potential to induce vascular fluid and protein extravasation.8 9 bFGF is a 16-kD monomeric peptide that differs functionally from VEGF in two important aspects: first, bFGF targets a wider variety of cell types in vitro, inducing the proliferation of SMCs and fibroblasts in addition to endothelial cells; second, bFGF is not known to alter vascular permeability.2
Recently, we found that intracoronary administration of bFGF10 or VEGF11 increases coronary collateral perfusion and myocardial vascular density in dogs with gradual coronary artery occlusion. We subsequently demonstrated that bFGF administered long-term (5 to 9 weeks) by the systemic arterial route is efficacious as well,12 although anemia and thrombocytopenia were observed in treated dogs. Interestingly, the most pronounced effect of bFGF was limited to an interval encompassing only 1 of the 9 weeks of the study.12 We therefore wanted to determine whether treatment limited to only 7 days might lead to maximal or near-maximal stimulation of collateral growth, an approach that would be more likely to avoid systemic side effects. Because bFGF is an SMC mitogen, we were concerned about its potential to accelerate atherosclerosis and the SMC proliferative response to vascular injury.13 14 15 16 17 Conversely, we thought it likely that VEGF, an endothelial cell–specific mitogen, would be less likely than bFGF to increase the SMC proliferative response to injury. Indeed, VEGF has recently been shown to enhance reendothelialization and decrease neointimal proliferation in a rat carotid injury model.18 Accordingly, the goals of this investigation were twofold: (1) to determine whether systemic administration of either bFGF or VEGF over a period of 7 days would enhance coronary collateral development and (2) to ascertain whether this beneficial effect could be achieved without accelerating injury-induced vascular neoin-timal formation. These goals were met in a canine model in which these two processes could be assessed concurrently.
The experimental protocol was approved by the Animal Care and Use Committee of the National Heart, Lung, and Blood Institute and was conducted in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals [Department of Health and Human Services Publication (NIH) 86-23, revised 1992] and NIH issuance 3040-2, Animal Care and Use in the Intramural Program. The animals used in this investigation were purchased specifically for this protocol and were used in no other study.
The study was conducted in two phases (Fig 1⇓). Phase 1 was a three-armed study designed to evaluate the effects of bFGF and VEGF on coronary collateral development and neointimal accumulation in response to iliofemoral injury. bFGF, VEGF, or saline was administered for 1 week; treatment was initiated 10 days after ameroid placement and on the day of vascular injury. On the basis of these results, phase 2 was a two-armed study (VEGF versus saline) to examine the effects of VEGF administered later in the course of collateral development; treatment was delivered 17 to 23 days after ameroid placement. Again, the test drug was begun on the day of vascular injury. The methods and personnel used in the two phases were identical.
Forty-three mongrel dogs of either sex were obtained from Haycock Kennels. Dogs were anesthetized with acepromazine 0.2 mg/kg IM, thiopental sodium 15 mg/kg IV, and inhaled methoxyflurane. A left thoracotomy was performed under sterile technique. An ameroid constrictor (Research Instruments and Manufacturing)19 was fitted on the proximal LCx before the takeoff of the first marginal branch, and a hydraulic balloon occluder was placed immediately distal to the ameroid (Fig 2⇓). Once implanted, the ameroid causes constriction over a period of 10 to 20 days, resulting in occlusion of the vessel. A Silastic catheter was positioned in the left atrial appendage for radiolabeled microsphere injections (for assessment of collateral blood flow) and drug administration. The pericardium and chest were closed, the tips of the balloon occluder and left atrial catheter were secured in the subcutaneous tissue of the back, and the dogs were allowed to recover. The left atrial catheter was flushed daily with saline to maintain patency.
Creation of Iliofemoral Arterial Injury
On the day of randomization to treatment (10 days after ameroid placement in phase 1 and 17 days after ameroid placement in phase 2), all dogs underwent unilateral iliofemoral balloon denudation injury. Under local lidocaine anesthesia and sterile technique, the right femoral artery was cannulated with a 7F sheath (Cordis Corp). A 5F Fogarty catheter was introduced through the sheath and advanced into the iliac artery to the 12.5-cm mark. The introducer was partially withdrawn to permit free movement of the Fogarty catheter. The balloon was inflated with 2.5 mL saline, and five passes back and forth were made to denude the endothelium. The balloon was then deflated to 2 mL and withdrawn to the 7.5-cm mark, into the more distal femoral artery, where the denudation procedure was repeated (the diameter of the femoral artery is less than that of the more proximal iliac artery). The catheter and introducer were removed, and hemostasis was achieved with direct pressure. This vessel was not subjected to arterial puncture or otherwise disturbed after injury until it was harvested for histological analysis 28 days later.
In phase 1, dogs were randomized to receive bFGF 1.74 mg (n=9), VEGF 0.72 mg (n=9), or saline (n=10) as a daily left atrial bolus during the interval 10 to 16 days after ameroid placement. In phase 2, dogs were randomized to receive VEGF 0.72 mg (n=6) or saline (n=5), with treatment 17 to 23 days after ameroid placement. In all dogs, iliofemoral injury was performed on the first day of treatment, and the peptide or saline was administered within 5 minutes of injury.
Selection of the doses was based on our previous studies of intracoronary bFGF10 and VEGF11 and the premise that 6% to 7% of cardiac output is distributed to the myocardium. Thus, the daily intracoronary doses of bFGF and VEGF used previously (110 and 45 μg, respectively)10 11 were multiplied by 1/0.065 to obtain the systemic doses used in this study.
Human recombinant VEGF (generously supplied by Genentech, Inc) 0.5 mg/mL, human recombinant bFGF (generously provided by Scios Nova) 10 mg/mL, and equivalent volumes of PBS were divided into aliquots, placed into coded vials, and stored at −20°C until used. All were diluted with 0.6 mL citrate buffer immediately before administration. Routine hematological and biochemical studies were performed weekly on all animals.
Hemodynamic Measurements and Quantification of Collateral Blood Flow
Regional myocardial blood flow was quantified by the reference sample technique20 on days 10 (pretreatment, phase 1), 17 (pretreatment, phase 2; posttreatment, phase 1), 24 (posttreatment, phase 2), and 38 (Fig 1⇑). All microsphere blood flow studies were performed with the dogs in the conscious state during maximal coronary vasodilatation. The dogs were lightly sedated with diazepam 1 to 2 mg/kg administered through the left atrial catheter. Under local lidocaine anesthesia, a 5F catheter (Cordis Corp) was inserted into the uninjured femoral artery for withdrawal of reference samples and measurement of arterial pressure with a strain-gauge manometer (P10EZ, Gould Electronics). The ECG and blood pressure were recorded continuously. Maximal coronary vasodilatation was induced by infusing chromonar 8 mg/kg (Intensain, Hoechst-Roussel Pharmaceuticals) into the left atrial catheter over a period of 30 minutes, as we have done previously.10 11 12 This drug elicits a fourfold to fivefold increase in myocardial blood flow without decreasing systemic blood pressure significantly.21 22 After infusion of chromonar, ≈3×106 radiolabeled microspheres, diameter 15 μm, were injected into the left atrial catheter as described previously.10 11 12 Because ameroid constrictors require 10 to 20 days to cause complete coronary occlusion, the LCx hydraulic occluder was temporarily inflated to arrest residual antegrade LCx perfusion during collateral blood flow measurements on days 10 and 17. The balloon occluder was permanently inflated on day 24 (after the blood flow measurement) to ensure complete LCx occlusion. On the final day (day 38 in phase 1, day 45 in phase 2; 28 days after iliofemoral injury in all animals), the dogs were heparinized (5000 U IV) and killed with an overdose of sodium pentobarbital and KCl, and the myocardium was perfusion-fixed with 10% buffered formaldehyde at physiological pressure.
Calculation of Myocardial Blood Flow
After fixation was complete, the heart was cut into 7-mm slices in the short-axis orientation. The left ventricular portions of the two central slices were divided into eight wedges and ultimately into endocardial and epicardial subwedges for microsphere blood flow analyses (Fig 2⇑). These 16 wedges (endocardial/epicardial pairs) were ranked with respect to flow, and the 4 wedges with the highest flow on day 10 were selected to represent the NZ; the 4 wedges with the most compromised flow were selected to represent the CZ, as we have done previously.10 11 12 Thus, endocardial and epicardial segments of 4 CZ and 4 NZ tissue wedges were used to compute mean CZ and NZ blood flows.
During maximal coronary vasodilatation, coronary blood flow is linearly related to perfusion pressure; therefore, collateral conductance was calculated as the quotient of maximal myocardial blood flow and mean arterial pressure (expressed in units of mL·min−1·100 g−1·mm Hg−1). Collateral flow was also expressed in relative terms as the ratio of CZ/NZ blood flow, as we have done previously.10 11 12
Evaluation of Infarct Size
Infarct size was calculated in phase 1 animals as we have done previously.10 11 The 16 wedges from the two central short-axis left ventricular slices were analyzed. After the samples were counted for microsphere blood flow analysis, they were embedded in paraffin and stained with Masson's trichrome. This method stains viable tissue red and collagen (or scar) blue. Slides were examined with a microscope (4× objective lens) and digitized in true color. A computer-controlled motorized stage was used to move the samples so that each slide was analyzed in its entirety. Blue-stained areas representing infarct were quantified by use of computer-assisted image analysis. By the same method, total tissue area was determined with a macro lens, and infarct size was expressed as a percentage: infarct size (%)=Σ infarct area (μm2)/Σ total area (μm2)×100%.
Histology of Noncardiac Tissue (Phase 1)
Tissue samples were removed from the liver, spleen, kidneys, lung, intestines, and skin. Tissues were fixed in formalin, embedded in paraffin, stained with hematoxylin-eosin, and subjected to histopathological examination by a veterinary pathologist blinded to treatment group. Rib samples were also examined from all dogs for bone marrow abnormalities.
Morphometric Analysis of Iliofemoral Arteries
The entire length of the iliac and femoral arteries was harvested and fixed by immersion in formaldehyde after the femoral end was tagged with a ligature. The arteries were sectioned at 1-mm intervals and embedded in paraffin in serial order. Sections 5 μm thick were cut from each block and stained with Movat pentachrome23 to delineate the internal elastic lamina. Sections were also stained for α-smooth muscle actin by immunohistochemical techniques to determine the phenotype of the predominant neointimal cell type.24 In all cases, the arterial section exhibiting the greatest degree of luminal narrowing was selected for planimetry. Percent stenosis was determined by computer-based image analysis. The percent luminal stenosis was calculated as the neointimal area divided by the area of the original lumen (the area circumscribed by the internal elastic lamina) times 100%. We previously determined that the cross-sectional area of the tunica media remains remarkably constant during the process of injury-induced neointimal proliferation25 ; therefore, the neointimal area was also normalized to the area of the tunica media, ie, the ratio of neointima to media. Lesion selection and planimetry were performed by two independent observers (D.F.L. and E.F.U.) unaware of treatment group. Interobserver variability was minimal (r=.94 for neointimal/medial ratio; r=1.00 for % stenosis). The methods and timing of treatment were identical in phases 1 and 2 (Fig 1⇑), and the results were combined in the analysis.
Acute Vasodilatory Effect of VEGF
Acute administration of bFGF has been reported to decrease arterial pressure in rats and rabbits,26 and we have made similar observations in dogs at the dose used in the present study.12 To assess potential hemodynamic effects of systemically administered VEGF, all dogs in phase 2 underwent hemodynamic monitoring during the left atrial injection of the treatment drug on day 17 (first day of treatment).
Pharmacokinetics of Systemic VEGF
In one animal, venous blood samples were obtained at multiple time points after injection of 0.72 mg VEGF into the left atrium. Samples were centrifuged at 4°C for 10 minutes, and serum was assayed for human VEGF with a solid-phase ELISA kit according to the manufacturer's instructions (catalog No. DVE00, R&D Systems, Inc).
Except where otherwise noted, data are expressed as mean±SEM. For the analysis of maximal collateral blood flow as a function of time, the measurement on day 10 was considered to constitute the baseline and was subtracted from the following three measurements. Using the terms time, treatment, time-by-treatment interaction, and dog within time, a repeated-measures analysis was used to examine the interaction between treatment and time in treated and control dogs. For blood chemistries and complete blood counts, a repeated-measures analysis was used, and P<.001 was considered significant. A one-way ANOVA was used to compare means of the three groups followed by individual two-tailed Student's t tests as appropriate.27
Thirty-nine of 44 dogs completed the studies. Four dogs died suddenly within 2 weeks of surgery, before randomization, presumably from sudden coronary occlusion or ameroid-induced coronary artery spasm. A fifth dog developed an extensive nonfatal myocardial infarction after surgery and was excluded from the study (before randomization).
Heart rate and blood pressure were similar in treated and control dogs throughout the course of the study (Table 1⇓).
Acute Hemodynamic Effects of VEGF Administration
Acute VEGF administration was associated with an immediate and prominent reduction in mean arterial pressure, with a peak effect 5 minutes after injection (mean arterial pressure, 92±4 mm Hg preinjection versus 64±8 mm Hg at t=5 minutes). At 10 minutes, the mean arterial pressure was 64±5 mm Hg (P=.008 preinjection versus 5 minutes; P=.004 preinjection versus 10 minutes). These changes were not associated with an increase in heart rate. Two animals underwent prolonged hemodynamic monitoring after VEGF injection. The hypotensive effect was half-maximal after 15 minutes, and arterial pressure returned to baseline after 90 minutes.
Pharmacokinetics of Systemic VEGF
After left atrial injection of 0.72 mg VEGF, there was a rapid distribution phase followed by an elimination phase with first-order kinetics (Fig 3⇓). The elimination half-life was ≈50 minutes.
Collateral Conductance and CZ/NZ Blood Flow Ratios: Phase 1
In the NZ, maximal coronary conductance was similar in the three treatment groups and relatively constant with respect to time (Table 2⇓). Maximal CZ conductance was similar in the three groups before randomization to treatment, and pretreatment CZ/NZ ratios were virtually identical in the three treatment groups. Maximal CZ conductance improved in all groups with respect to time (Table 2⇓, Fig 4⇓). At the first posttreatment time point, 7 days after treatment started (day 17), maximal CZ conductance in bFGF-treated dogs had increased such that conductance was 68% higher than that of controls (P<.01). This increase was also reflected in the CZ/NZ ratio (P<.01, bFGF versus control; P<.003, bFGF versus VEGF; Table 2⇓, Fig 5⇓). This pattern persisted at subsequent time points after treatment was discontinued: administration of bFGF was associated with 49% and 46% increases in maximal CZ conductance on days 24 and 38, respectively (P<.01 versus control for both time points), with corresponding and significant increases in the CZ/NZ ratio. CZ conductance was also increased in bFGF-treated dogs relative to VEGF-treated dogs (P<.03 for days 24 and 38), and a significant difference in CZ/NZ ratio was demonstrable on days 17 and 38 (P<.03, bFGF versus VEGF). Similar differences were noted between the bFGF and control groups when endocardial and epicardial flows were analyzed separately, indicating that the bFGF-induced improvement in maximal collateral flow was not restricted to either the endocardium or epicardium.
Collateral Conductance and CZ/NZ Blood Flow Ratios: Phase 2
As expected, maximal CZ conductance and the CZ/NZ ratios improved in both groups with respect to time. As in phase 1, there were no significant differences in maximal CZ conductance or CZ/NZ ratios between VEGF-treated and control dogs throughout the study (Fig 6⇓). Final CZ/NZ ratios were virtually identical in VEGF-treated and control dogs (mean, 0.41) and were similar to the final CZ/NZ ratios in control and VEGF-treated dogs in phase 1 (mean, 0.43 for both).
Infarct size in phase 1 dogs was insignificant, averaging 1.71±0.87%, 0.24±0.10%, and 1.17±0.93% of the two midequatorial left ventricular slices in control dogs, bFGF-treated dogs, and VEGF-treated dogs, respectively (ANOVA, P=.39). Infarct size was not calculated in phase 2 dogs. In none of the phase 2 animals was a large infarct apparent during examination of the gross pathology.
Effects of bFGF and VEGF on the Vascular Response to Injury
Arterial sections from two dogs had evidence of extensive luminal thrombus and were excluded. Both dogs had received VEGF (χ2 test, P=NS). An additional dog was excluded because of technical difficulty in creating the lesion. Neointimal response to injury was observed within the iliofemoral arteries in all three groups (Figs 7⇓ and 8).⇓ Percent luminal stenosis (phases 1 and 2 combined) was 22±3%, 25±5%, and 50±8% in controls, bFGF-treated, and VEGF-treated dogs, respectively (ANOVA, P=.0025; bFGF versus control, P=NS; VEGF versus control, P=.0027; VEGF versus bFGF, P=.035). Neointimal/medial ratios were 0.33±0.05, 0.45±0.10, and 0.70±0.13 in controls, bFGF-treated, and VEGF-treated dogs, respectively (ANOVA, P=.021; bFGF versus control, P=.23; VEGF versus control, P=.009; VEGF versus bFGF, P=.20). Inflammation was not observed in any of the arterial sections. Immunohistochemical staining for α-smooth muscle actin identified the SMC as the predominant cell type in the neointima.
Biochemical and Hematological Data: Phase 1
Serum creatine phosphokinase, glucose, urea nitrogen, creatinine, albumin, SGOT, SGPT, alkaline phosphatase, bilirubin, calcium, and phosphate were similar in all three groups throughout the course of the study (Table 3⇓). In bFGF-treated dogs, there was a reversible decrease in the circulating platelet count, with the platelet count decreasing from 304±39×103/mm3 (pretreatment) to 143±34×103/mm3 by the conclusion of treatment (P<.01). One week after treatment was suspended, the platelet count recovered to 325±39×103/mm3, and it remained at this level during the final 2 weeks. The platelet count was consistent across time in controls and VEGF-treated dogs, and in no group was the hematocrit or total leukocyte count significantly affected by treatment.
Histopathology of Noncardiac Tissues: Phase 1
All animals exhibited a minimal degree of chronic interstitial nephritis, presumably a reaction to microspheres deposited in the renal parenchyma. There was no difference in severity between groups. Other tissues were similar in all groups.
In previous investigations, we demonstrated that administration of either bFGF10 12 or VEGF11 for ≥4 weeks substantially increased maximal collateral blood flow and vascular density in dogs with chronic coronary occlusion. Because of the potential for untoward side effects when potent mitogens are administered over an extended time period, the present investigation was undertaken to determine whether a far briefer (1-week) treatment period would suffice to effectively increase coronary collateral development. We also sought to determine whether such a beneficial effect could be achieved without exacerbating the neointimal response to vascular injury, one of the processes believed to play a critical role in the pathogenesis of atherosclerosis, restenosis after angioplasty,13 and the precipitation of unstable angina.28 Thus, administration of angiogenic peptides could have a dual effect in patients with occlusive coronary disease: promoting collateral formation but simultaneously inducing neointimal SMC accumulation and ultimately exacerbating atherogenic processes. We were particularly interested in testing the concept that the net effect of VEGF, a relatively specific endothelial cell mitogen, would be more likely to favor angiogenesis rather than SMC proliferation in response to injury compared with that of bFGF, which induces proliferation in a much more diverse population of cell types.
Effect of bFGF on Maximal Coronary Collateral Flow
In dogs with single-vessel coronary occlusion and myocardial ischemia, bFGF administration (1.74 mg/d for 7 days) was associated with a 68% increase in maximal coronary collateral perfusion at the end of treatment (day 17). This improvement was sustained such that maximal collateral blood flow measured 3 weeks after cessation of treatment was increased by 46% relative to controls. The success of the much-abbreviated treatment interval was not unanticipated. In the studies in which bFGF was administered for 4 to 9 weeks,10 12 we observed that the bFGF-induced improvement in maximal collateral flow occurred predominantly during a 1-week interval (10 to 17 days after ameroid placement), with collateral flow of the treated and control groups moving in parallel thereafter. This observation led us to speculate that treatment limited to 7 days might constitute an adequate stimulus to enhance collateral flow. The results of the present investigation confirm the validity of this supposition and represent an important extension of our previous studies10 12 : that prolongation of treatment beyond 7 days is unnecessary.
The mechanism of action of bFGF on collateral development in this model may be twofold. Presumably bFGF, a potent inducer of vascular cell proliferation and migration, enhanced collateral growth through direct angiogenic effects. bFGF has also been reported to be an endothelium-dependent vasodilator,26 and we have observed that pharmacological doses of bFGF cause both coronary10 and systemic12 vasodilatation. Vasodilators such as dipyridamole have been observed to increase myocardial capillary density or collateral blood flow.29 Thus, bFGF-induced vasodilatation may have played a secondary role in promoting collateral formation.
Effect of VEGF on Maximal Coronary Collateral Flow
In contradistinction to what was observed with bFGF and at variance with data we reported previously,11 systemic arterial administration of VEGF failed to increase maximal collateral flow in this investigation. This was substantiated by two independent segments of the present study (phases 1 and 2). The lack of effect could be due to any of several differences between the two studies: the VEGF dose, route of administration, and timing and/or duration of treatment.
VEGF Dose and Route of Administration
In our previous study, in which VEGF was found to enhance coronary collateral development,11 we gave VEGF 45 μg IC directly into the vessel supplying the ischemic myocardium, and this dose was administered 20 times over a 4-week period. In the present study, a 720-μg dose of VEGF was administered into the systemic arterial circulation on seven occasions. The systemic arterial doses of both VEGF and bFGF were based on our previous intracoronary studies,10 11 in which the peptides were found to be equally effective in promoting collateral development. For the present study, the dose of each peptide was increased by a factor of 1/0.065 to compensate for the fact that ≈6% to 7% of cardiac output is distributed to the myocardium. Theoretically, this dose adjustment would ensure equivalent VEGF concentrations in myocardial tissue at first pass through the coronary circulation. This would not, however, expose the coronary vasculature to an equivalent concentration of VEGF. Considering that VEGF binds to specific receptors in vascular endothelial cells and proteoglycans in the extracellular matrix, the intravascular concentration may be more important for biological effect than the tissue concentration. In our previous study,11 it is reasonable to assume that at the point of entry, 45 μg VEGF was distributed in a volume of ≈4 to 5 mL (the volume of the distal LCx and its main tributaries), such that the coronary vascular endothelium would have been exposed very briefly to a VEGF concentration of ≈10 μg/mL. In the present study, injection of 720 μg VEGF into a systemic blood volume of 2 L would provide an initial systemic concentration of only 0.36 μg/mL. Thus, the initial intracoronary VEGF concentration in the present study was ≈1.5 orders of magnitude less than that in the previous intracoronary study.11 Furthermore, VEGF was administered seven times in the present study, whereas it was given 20 times in the latter study. Thus, a substantial reduction in the cumulative coronary vascular exposure to VEGF in the present investigation provides a plausible explanation for the conflicting results of the two studies.
It must also be noted that VEGF administration was associated with a profound decrease in mean arterial pressure (to 64 mm Hg), most likely due to endothelium-dependent relaxation.30 Thus, it is unlikely that the systemic VEGF dose could have been increased substantially. Moreover, the VEGF dose used in this study caused deleterious effects on injured vessels: a higher dose would be expected to cause further neointimal accumulation (see below).
Timing and Duration of Treatment
For the initial study (phase 1), data accrued previously when bFGF was used in this model provided the rationale for the selection of the treatment interval (10 to 16 days after ameroid placement for both bFGF and VEGF).10 12 Although this strategy worked for bFGF, VEGF did not enhance maximal collateral flow. With these data at hand and after careful review of our previous VEGF data,11 we decided to repeat the VEGF and control arms of the study with a modified treatment schedule. In considering our intracoronary VEGF data in retrospect,11 it was apparent that there was little improvement in collateral flow 10 to 16 days after ameroid placement; the major effect of VEGF became apparent after the second week of treatment (days 17 to 23). Thus, phase 2 was designed to determine whether VEGF could enhance collateral development if treatment were delayed by 1 week relative to the ischemic stimulus. This approach also failed to improve maximal collateral perfusion. To enhance coronary collateral development in this model, it appears that VEGF must be administered by the intracoronary route, and/or an extended duration of peptide delivery is essential.
Recently, a single dose of intra-arterial31 or intravenous32 VEGF has been shown to enhance collateral vessel formation in a rabbit ischemic hindlimb model. These data are promising and suggest that there may be model-specific differences between the ischemic rabbit hindlimb and the canine myocardium.
Effects of bFGF and VEGF on the Vascular Response to Injury
We had been concerned that the beneficial effects of bFGF on collateral development might be accompanied by an acceleration of processes associated with atherogenesis. In fact, bFGF did not significantly increase neointimal SMC accumulation after vascular injury, even though the dose administered achieved a substantial increase in maximal coronary collateral flow. These data are encouraging; however, some caution must be observed in their interpretation because the study lacks the statistical power to prove a negative result. Nevertheless, these observations do suggest that the salutary angiogenic effects of bFGF can be dissociated from potential deleterious responses to vascular injury. A critical component of this dissociation may be the brief period of bFGF treatment. It is also plausible that bFGF, by inducing thrombocytopenia, decreased the availability of platelet-derived growth factor and/or other platelet-associated mitogens, thereby counterbalancing the mitogenic effects of bFGF on SMCs such that there was no net effect on neointimal thickening. Even though this theory is attractive, we found no relation between platelet count and the degree of neointimal accumulation in individual dogs to support this theory.
Although VEGF failed to produce an angiogenic effect in this investigation, it did exert a potent biological effect on neointimal expansion after vascular injury (Fig 7⇑). This has important ramifications as to its potential role as an agent to promote collateral development, because it is likely that increasing the dose or duration of treatment would be associated with an even greater propensity to exacerbate SMC accumulation in injured vessels. To the best of our knowledge, this is the first study to demonstrate the intimal SMC proliferative effect of VEGF in vivo. Indeed, VEGF has been shown recently to enhance reendothelialization and decrease neointimal proliferation in a rat carotid injury model.18 Thus, the VEGF-induced increase in neointimal development observed in this experimental model was unprecedented and unexpected, because the mitogenic effect of VEGF is highly specific for endothelial cells in vitro, and neointimal hyperplasia is generally a phenomenon of SMC proliferation.33 34
It is plausible that the apparent discordance between the endothelial cell specificity of VEGF in vitro and the neointimal thickening observed in vivo is related to the potent effects of VEGF on vascular permeability.3 8 9 A VEGF-mediated increase in vascular permeability could result in leakage of various growth factors at the site of vascular injury, such as platelet-derived growth factor and thrombin. Moreover, VEGF is a direct chemoattractant for monocytes,35 and the monocyte/macrophage has been implicated in SMC proliferation after arterial injury.36 These theories are purely speculative, however, and the present study was not designed to evaluate these potential causalities. Nonetheless, it is apparent that VEGF may cause complex responses in vivo that would not be predicted on the basis of its known endothelial cell specificity. These findings have important ramifications if VEGF is considered for systemic administration to promote collateral growth.
It should be noted that the neointimal response to vascular injury was observed in normal arteries subjected to balloon denudation injury. Although the histological appearance of the vessels bore a striking resemblance to the characteristic lesion of restenosis after angioplasty, the use of normal arteries in normocholesterolemic animals is an inherent limitation of this study. Our results may not be predictive of the response within the complex biological milieu of an atherosclerotic lesion or relevant to an atherosclerotic artery after angioplasty. In light of the limitations of this study, the divergent results obtained by others,14 15 18 and the fact that angiogenic growth factors are entering into clinical trials,37 we believe that further studies in this area are warranted.
Adverse Systemic Effects
The kidney and bone marrow have been identified previously as the main targets of bFGF toxicity.38 39 40 Mazue´ et al40 observed renal toxicity in cynomolgus monkeys receiving intravenous bFGF 100 μg·kg−1·d−1 for 7 days, a dose similar to that used in the present study. Moderate reversible thrombocytopenia was observed in bFGF-treated animals, without evidence of hemorrhage or bruising. There were no other apparent hematological abnormalities and no detectable alterations in renal function tests, electrolytes, or hepatic transaminases. The short duration of treatment (7 days) in the present investigation may have served to minimize systemic side effects, a finding that has important clinical implications. We did not observe (by light microscopy) specific pathological findings attributable to bFGF; however, the kidneys were not subjected to electron microscopy, and proteinuria was not assessed. The bone marrow was not analyzed until 21 days after the conclusion of treatment; therefore, subtle or reversible pathological changes may have been overlooked by our analyses.
The potential systemic toxicity of VEGF has not been characterized previously. We did not observe hematological or biochemical evidence of VEGF toxicity. Moreover, no pathological changes were attributable to VEGF by our light-microscopic analyses.
The results of the present study confirm the feasibility of effective pharmacological enhancement of coronary collateral development with abbreviated exposure to exogenous bFGF, an intervention that appears to be relatively safe. Systemic arterial VEGF, administered at a dose sufficient to induce significant arterial hypotension, failed to improve maximal myocardial collateral blood flow. These results were verified in two independent phases of this investigation, in which two different treatment schedules were evaluated. VEGF, but not bFGF, was found to exacerbate the neointimal response to vascular injury, an effect that would not be predicted on the basis of the endothelial cell specificity of VEGF. These data provide support for the clinical investigation of bFGF as an agent to promote collateral development in selected patients with ischemic heart disease. In future clinical trials in which angiogenic peptides are targeted to induce collateral development, evidence of vascular neointimal thickening should be sought as an important complication of such a therapeutic strategy.
Selected Abbreviations and Acronyms
|bFGF||=||basic fibroblast growth factor|
|LCx||=||left circumflex coronary artery|
|SMC||=||smooth muscle cell|
|VEGF||=||vascular endothelial growth factor|
The authors thank V. Hampshire, J. Bacher, and their staffs for providing veterinary care, W. Ballew for clinical laboratory support, Genentech, Inc, for providing VEGF, Scios Nova, Inc, for supplying bFGF, and Hoechst-Roussel Pharmaceuticals for providing chromonar.
Presented in part at the 43rd Annual Scientific Sessions of the American College of Cardiology, Atlanta, Ga, March 13-16, 1994.
- Received September 18, 1995.
- Revision received March 7, 1996.
- Accepted March 13, 1996.
- Copyright © 1996 by American Heart Association
Connolly DT, Heuvelman DM, Nelson R, Olander JV, Eppley BL, Delfino JJ, Siegel NR, Leimgruber RM, Feder J. Tumor vascular permeability factor stimulates endothelial cell growth and angiogenesis. J Clin Invest. 1989;84:1470-1478.
Gospodarowicz D, Abraham JA, Schilling J. Isolation and characterization of a vascular endothelial cell mitogen produced by pituitary-derived folliculo stellate cells. Proc Natl Acad Sci U S A. 1989;86:7311-7315.
Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science. 1989;246:1306-1309.
Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey VS, Dvorak HF. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science. 1983;219:983-985.
Keck PJ, Hauser SD, Krivi G, Sanzo K, Warren T, Feder J, Connolly DT. Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science. 1989;246:1309-1312.
Unger EF, Banai S, Shou M, Lazarous DF, Jaklitsch MT, Scheinowitz M, Correa R, Klingbeil C, Epstein SE. Basic fibroblast growth factor enhances myocardial collateral flow in a canine model. Am J Physiol. 1994;266:H1588-H1595.
Banai S, Jaklitsch MT, Shou M, Lazarous DF, Scheinowitz M, Biro S, Epstein SE, Unger EF. Angiogenic-induced enhancement of collateral blood flow to ischemic myocardium by vascular endothelial growth factor in dogs. Circulation. 1994;89:2183-2189.
Lazarous DF, Scheinowitz M, Shou M, Hodge E, Rajanayagam MAS, Hunsberger S, Robison WG, Stiber JA, Correa R, Epstein SE, Unger EF. Effects of chronic systemic administration of basic fibroblast growth factor on collateral development in the canine heart. Circulation. 1995;91:145-153.
Lindner V, Lappi DA, Baird A, Majack RA, Reidy MA. Role of basic fibroblast growth factor in vascular lesion formation. Circ Res. 1991;68:106-113.
Edelman ER, Nugent MA, Smith LT, Karnovsky MJ. Basic fibroblast growth factor enhances the coupling of intimal hyperplasia and proliferation of vasa vasorum in injured rat arteries. J Clin Invest. 1992;89:465-473.
Banai S, Jaklitsch MT, Casscells W, Shou M, Shrivastav S, Correa R, Epstein SE, Unger EF. Effects of acidic fibroblast growth factor on normal and ischemic myocardium. Circ Res. 1991;69:76-85.
Asahara T, Bauters C, Pastore C, Kearney M, Rossow S, Bunting S, Ferrara N, Symes JF, Isner JM. Local delivery of vascular endothelial growth factor accelerates reendothelialization and attenuates intimal hyperplasia in balloon-injured rat carotid artery. Circulation. 1995;91:2793-2801.
Baer RW, Payne BD, Verrier ED, Vlahakes GJ, Molodowitch D, Uhlig PN, Hoffman JIE. Increased number of myocardial blood flow measurements with radionuclide-labeled microspheres. Am J Physiol. 1984;246:H418-H434.
Movat H. Demonstration of all connective tissue elements in a single section. Arch Pathol Lab Med. 1955;60:289-295.
Gabbiani G, Kocher O, Bloom WS. Actin expression in smooth muscle cells of rat aortic intimal thickening, human atheromatous plaque, and cultured rat aortic media. J Clin Invest. 1984;74:148-152.
Banai S, Shou M, Correa R, Jaklitsch MT, Epstein SE, Unger EF. The rabbit ear model of injury-induced arterial smooth muscle cell proliferation: kinetics, reproducibility, and implications. Circ Res. 1991;69:748-756.
Cuevas P, Carceller F, Ortega S, Zazo M, Nieto I, Gime´nez-Gallego G. Hypotensive activity of fibroblast growth factor. Science. 1991;254:1208-1210.
Dowdy S, Wearden S. Statistics for Research. New York, NY: John Wiley & Sons, Inc; 1983.
Flugelman MY, Virmani R, Correa R, Yu Z-Y, Farb A, Leon MB, Elami A, Fu Y-M, Casscells W, Epstein SE. Smooth muscle cell abundance and fibroblast growth factors in coronary lesions of patients with nonfatal unstable angina: a clue to the mechanism of transformation from the stable to the unstable clinical state. Circulation. 1993;88:2493-2500.
Cohen MV. Coronary Collaterals: Clinical and Experimental Observations. Mount Kisco, NY: Futura; 1985:419-446.
Ku DD, Zaleski JK, Liu S, Brock TA. Vascular endothelial growth factor induces EDRF-dependent relaxation in coronary arteries. Am J Physiol. 1993;265:H586-H592.
Takeshita S, Zheng LP, Brogi E, Kearney M, Pu LQ, Bunting S, Ferrara N, Symes JF, Isner JM. Therapeutic angiogenesis: a single intraarterial bolus of vascular endothelial growth factor augments revascularization in a rabbit ischemic hind limb model. J Clin Invest. 1994;93:662-670.
Clauss M, Gerlach M, Gerlach H, Brett J, Wang F, Familletti PC, Pan Y-CE, Olander JV, Connolly DT, Stern D. Vascular permeability factor: a tumor-derived polypeptide that induces endothelial cell and monocyte procoagulant activity, and promotes monocyte migration. J Exp Med. 1990;172:1535-1545.
Isner JM, Walsh K, Symes J, Pieczek A, Takeshita S, Lowry J, Rossow S, Rosenfield K, Weir L, Brogi E, Schainfeld R. Arterial gene therapy for therapeutic angiogenesis in patients with peripheral artery disease. Circulation. 1995;91:2687-2692.
Mazue´ G, Bertolero F, Garofano L, Brughera M, Carminati P. Experience with the preclinical assessment of basic fibroblast growth factor (bFGF). Toxicol Lett. 1992;64-65:329-338.
Mazue´ G, Newman AJ, Scampini G, Torre PD, Hard GC, Iatropoulos MJ, Williams GM, Bagnasco SM. The histopathology of kidney changes in rats and monkeys following intravenous administration of massive doses of FCE 26184, human basic fibroblast growth factor. Toxicol Pathol. 1993;21:490-501.