- Email: [email protected]
Intrapericardial basic fibroblast growth factor induces myocardial angiogenesis in a rabbit model of chronic ischemia
Intrapericardial basic fibroblast growth factor induces myocardial angiogenesis in a rabbit model of chronic ischemia The objective of this study was to determine whether basic fibroblast growth factor (bFGF), a known angiogenic factor, can promote new vessel growth when infused within the pericardial space in a model of chronic myocardial ischemia. Intravenous angiotensin II (All) was infused to induce left ventricular hypertrophy and concomitant ischemia in New Zealand white rabbits. Basic FGF was infused into the intrapericardial space with an osmotic pump, Animals were assigned to one of four groups: group 1 received intrapericardial bFGF and intravenous All, group 2 received intrapericardial bFGF and intravenous saline solution, group 3 received intrapericardial albumin and intravenous All, and group 4 received intravenous All only. Epicardial angiogenesis was graded histologically on a scale of 0 to 2. Animals receiving intravenous administration of All displayed left ventricular hypertrophy that disproportionately affected the interventricular septum with a wall thickness of 5.62 _+ 1.00 mm versus 3.98 ± 0.61 mm in the All group and the saline solution control group, respectively (p < 0.005). A highly localized angiogenic effect of bFGF was observed. The mean angiogenesis scores were 1.9, 1.4, 1.3, and 0.2 (p < 0.001) with an angiogenesis score of 2 (marked increase in vascularity) noted in 86%, 40%, 43%, and 0% of hearts in groups 1 through 4, respectively. We conclude that intrapericardial bFGF enhances new epicardial small-vessel growth in a rabbit model; furthermore this effect is enhanced in the presence of left ventricular hypertrophy. (AM HEARTJ 1995;129:924-31.)
Charles Landau, MD, Alice K. Jacobs, MD, and Christian C. Haudenschild, MD Boston, Mass.
Compromised microvascular circulation is an important component of several cardiovascular diseases including syndromes of acute coronary ischemia, 1 hypertensive heart disease with left ventricular hypertrophy, 2 chronic congestive heart failure, 3 and diabetic peripheral vascular insufficiency. 4 In each of these conditions an unfavorable or inappropriate ratio of capillary density to tissue oxygen demand appears to contribute to the extent, severity, or persistence of muscle dysfunction. 36 In theory interventions to increase capillary density in underperfused tissue beds could have a meaningful palliative effect Prom the Evans Memorial Department of Clinical Research and the Section of Cardiology, Department of Medicine, Boston University Medical Center (CL, AKJ), and the Mallory Institute of Pathology (CCH), Boston. Supported by a Biomedical General Research Support Grant of University Hospital, Boston. Received for publication Aug. 25, 1994; accepted Oct. 7, 1994. Reprint requests: Charles Landau, MD, University of Texas Southwestern Medical Center, Cardiology Division, 5323 Harry Hines Blvd., Dallas, TX 75235-9047. Copyright © 1995 by Mosby-Year Book, Inc. 0002-8703/95/$3.00 + 0
924
4/1/61996
in these conditions. This possibility is supported by several observations including (1) a correlation between improvement in functional class and skeletal muscle capillary density in patients with chronic congestive heart failure subjected to graded exercise programs, 7 (2) an improvement in postinfarction left ventricular systolic function in experimental animals as a result of myocardial angiogenesis induced by administration of exogenous basic fibroblast growth factor (bFGF), an angiogenic peptide, 5 and (3) decreased skeletal muscle ischemia in animals receiving local administration of bFGF in a rabbit hind limb model of vascular insufficiency. 8, 9 Basic fibroblast growth factor (bFGF) is a 17 kd cationic polypeptide consisting of 154 amino acids. Basic FGF is mitogenic for a number of primary and established cell types in culture including vascular endothelial cells. 1° Although the molecular mechanism by which bFGF exerts its mitogenic effect is incompletely defined, cells exposed to bFGF demonstrate increased expression of several cellular protooncogenes including c-myc. 11 Basic FGF also has
Volume 129, Number 5 American Heart Journal
Landau, Jacobs, and Haudenschild 925 New Zealand White Rabbits
Implan
.
.
.
Implantation of angiotensin 1I (All) osmotic pump into external jugular vein
Intrapericard aI(IP)bFGF
1-4 w e e k s
Grou ~1 IP bFGF+IV All n=7
.
Implantation of saline osmotic p u m p into external jugular vein
Intraperic~ rdial(IP)bFGF
Intraperk rdial(IP) Albumin
1-4 weeks
1-4 weeks
Grou ~2 IP bFGF+W Saline n=5
Implantation of angiotensin I1 (All) osmotic p u m p into external jugular vein
Grou, 3 IP Alb + IV AlI n=7
1-4 weeks
Gro~, 4 1V AII only n=5
Sacrifice,Neovascular response score
Fig. 1. Flow diagram of animal protocol. IP, Intrapericardial; IV, intravenous; AII, angiotensin II; bFGF, basic fibroblast growth factor; Alb, albumin.
pleiotropic effects on the promotion of capillary growth and invasion. 1°13 It has been shown to induce angiogenesis in several animal models. Applied directly or adsorbed to implantable synthetic materials that release the peptide over time, bFGF has been demonstrated to induce angiogenesis in chick chorioallantoic membrane, ~° rabbit cornea, ~° rat peritoneal cavity, 14,15 rat neck, 14 rabbit mandible, 16 rat perinephric capsule, 17 and in the adventitia or media of rat carotid arteries. 12 In each of these experiments prolonged exposure to bFGF (at least 1 week) appeared to be required for an angiogenic response. Neovascularization has also been described in skeletal and cardiac muscle exposed to bFGF in the setting of ischemia. 5, 8, 9,18 In theory locally applied angiogenic factors could potentially enhance the efficacy of surgical or percutaneous revascularization procedures by promoting new vessel growth in ischemic regions. In a canine model of myocardial ischemia and infarction, acidic fibroblast growth factor (aFGF) applied to the epi-
cardial surface did not demonstrate evidence of new capillary growth, 19 but the effects o f bFGF in this setting have not been evaluated previously. Accordingly this study was undertaken to assess the ability of bFGF to induce angiogenesis when applied to the pericardial space under conditions of increased myocardial oxygen demand. METHODS
All procedures were performed in male New Zealand White rabbits under the guidelines of the Animal Welfare Act. The protocol was approved by the Institutional Animal Care and Use Committee of the Boston University Medical Center. Animals received continuous infusion osmotic pumps that delivered intravenous solutions, intrapericardial solutions, or both for periods of up to 28 days. It is well established that acquired left ventricular hypertrophy in adult animals is associated with inadequate compensatory capillary growth. 2°,21Intravenous angiotensin II (AII) was used to rapidly induce left ventricular hypertrophy,2 which provided an ischemic milieu caused by the inadequate density of capillaries present to supply the
May 1995
926
Landau, Jacobs, and Haudenschild
additional myocardium. 22 Rabbits were assigned to one of four groups: group 1 received intrapericardial bFGF and intravenous AII, group 2 received intrapericardial bFGF and intravenous saline solution, group 3 received intrapericardial albumin and intravenous AII, and group 4 received intravenous AII only. Tail cuff blood pressures were measured twice weekly for the duration of the experiment. A flow diagram (Fig. 1) summarizes the protocol sequence. Intrapericardial catheter placement, New Zealand White rabbits weighing 2.0 to 4.3 kg were quarantined for a period of 1 week to exclude communicable infections. Because pilot studies revealed a 10 % mortality rate after pericardial instrumentation, the initial procedure in groups 1 through 3 was implantation of a pericardial catheter to create a portal for intrapericardial infusions. To achieve this goal animals were anesthetized with xylazine (5 mg/kg) and ketamine (35 mg/kg) given subcutaneously. The subxiphoid region was shaved and prepped in sterile fashion. The animal was placed in the supine position and was rotated slightly to the left so that the heart rested to the left of midline under fluoroscopy. Lidocaine (Xylocaine) (2 %) was infused subcutaneously, and a longitudinal 3 cm incision was made in the midline with the superior aspect overlying the xiphoid process. Blunt dissection was undertaken, forming a tunnel through superficial layers. A 22-gauge spinal needle placed within a 5F dilator was inserted through the tunnel. The proximal end of the needle was connected to a tuohy adapter through which a 0.014-inch diameter Flex guide wire (USCI, Billerica, Mass.) was placed. The guide wire was positioned so that the distal end was 1 cm proximal to the needle tip. A 10 ml syringe filled with 70 % diatrizoate Renografin-76, Squibb, Princeton, N.J.) was attached to the infusion port of the tuohy adapter. The needle assembly was then directed at a 15-degree angle to the skin surface and was advanced under fluoroscopy towards the cardiac silhouette to the left of the midline. It was advanced gently until it popped through the diaphragm into the pericardial space. Correct needle position was confirmed by pulsatile motion of the needle tip. Contrast was injected to assess the location of the distal needle. An intraventricular location was indicated by rapid dye clearance, whereas an intrapericardial location of the needle was indicated by gravity-dependent pooling of contrast in the pericardial space. Dye staining represented an intramyocardial or intrapleural location. If necessary, the needle was repositioned with repeated contrast injections until a pericardial location was achieved. The guide wire was advanced into the pericardial space as confirmed by the wire remaining within the cardiac silhouette. The 5F dilator was passed over the wire, and the needle and wire were removed. The location of the dilator was confirmed with a contrast injection. Next a 0.018-inch diameter Flex guide wire (USCI) was passed through the dilator and was looped superiorly and then inferiorly around the heart so that the distal end rested in the region of the anterolateral wall of the left ventricle. The dilator was removed, and a 20 cm length of polyethylene tubing (PE-60, Fisher Scientific, Santa Clara,
AmericanHeartJournal
Calif.) was advanced over the wire to the distal end of the wire. After the wire was removed, position was confirmed by contrast injection through the tubing. Saline solution was flushed through the catheter, and it was anchored to abdominal muscle fascia with 4-0 silk suture. The proximal end of the catheter was placed in a subcutaneous pocket, and the skin was closed with absorbable suture. Animals received 100,000 U intramuscular penicillin on the day of the procedure and on alternate days for a total of three doses. Neck pump placement. Two days after intrapericardial catheter placement was done, osmotic pumps (Alza Model 2ML4, Alza Corp, Palo Alto, Calif.) were loaded in sterile fashion with the appropriate dose of angiotensin II (AII) or normal saline solution (Fig. 1) and were attached to a 5 cm length of PE-60 tubing. The pumps were incubated at 37 ° C for a minimum of 4 hours to activate infusion activity as specified by the manufacturer. Animals were anesthetized as described previously. The neck region was shaved and draped in sterile fashion with 2 % Xylocaine instilled subcutaneously. A midline 3 to 4 cm neck incision was made, the external jugular vein was isolated, and a subcutaneous pocket was formed. The distal aspect of the vein was tied off, and a venotomy was made. The pump was removed from the heated bath and positioned in the pocket. The attached tubing was trimmed so that approximately 2 cm of catheter could be advanced into the vein. An intravenous dose of 200 U heparin was given, the tubing was secured into its intravenous location with silk ties, and the pocket and skin were sutured closed. Antibiotics were administered as indicated previously. Because of the dehydrating effects of the AII, all animals with AII pumps were weighed three times per week and received 5 % dextrose in 0.9 % normal saline solution at a rate of 30 ml/day by subcutaneous infusion. If weight loss was evident, this administration was increased to 50 ml/day, until animal weight was stabilized. After I week of intravenous AII or saline solution infusion, the pericardial pump was attached in animals in groups 1 through 3 to the previously placed catheter as described in the following section. Pericardial pump attachment. Pericardial pumps (Alza, model 2ML4) were loaded in sterile fashion with 5/~g basic human recombinant FGF (Gibco, Grand Island, N.Y.) diluted to a volume of 2 ml with 0.1% bovine serum albumin (BSA) in phosphate-buffered saline solution filtered through a 0.45 ~m millipore filter. The resultant dose of bFGF was 180 ng/day. It has been demonstrated that bFGF retains its biologic activity at 37 ° C for periods of up to 4 weeks when dispensed from a sequestered source. 23Animal pumps in the control group were loaded with 0.1% BSA only (Fig. 1). The rabbit was anesthetized as described previously. The previously placed subxiphoid sutures were removed, and the wound was partially opened. The pericardial catheter was identified and isolated, and its position in the pericardial space was reconfirmed with a contrast injection. A subcutaneous pocket was created with blunt dissection.
Volume 129, Number 5 AmericanHeartJournal
927
Landau, Jacobs, and Haudenschild
Table I. Baseline characteristics
AII dose (~g/kg/hr) Days of AII infusion Days of IP infusion Blood pressure (mm Hg)
Group 1 (IP bFGF + I V AII, n = 7)
Group 2 (IP bFGF + I V saline solution, n = 5)
Group 3 (IP Alb + I V AII, n = 7)
Group 4 (IV AII, n = 5)
p Value
17.8 + 6.8 24.3 _+ 6.8 22.7 + 9.2 130 ± 20
NA NA 28 _+ 0 113 __ 12
18.7 + 8.6 20.9 _+ 7.9 18 + 9.9 131 ± 16
19.8 ± 12.1 16.6 ± 7.3 NA 109 _+ 26
NS NS NS NS
IP, Intrapericardial; IV, intravenous; Alb, albumin; A/I, angiotensin II; bFGF, basic fibroblast growth factor; NS, not significant; NA, not applicable.
The p u m p was placed in the pocket and was attached to the pericardial tubing. The tubing was anchored to the pump with silk ties, and the pump was then sutured to subcutaneous fascia. The skin was closed, and antibiotics were administered as previously described. Two weeks after the p u m p was implanted, the animal was again anesthetized, and the pump was exposed to confirm that it was still adequately positioned in the pocket and was connected to the pericardial tubing. Euthanization. Four weeks after pericardial pump placem e n t the animal was again anesthetized. A midline incision from the sternal notch to the abdominal region adjacent to the pericardial pump was made. Pentothal was given in a lethal intravenous dose of 25 mg/kg. The heart was exposed by a median sternotomy, and the pericardial catheter was carefully located. Its proximal portion was ligated in the supradiaphragmatic region so that its terminal position in the pericardial space was preserved. The heart and lungs were quickly removed, and the heart was perfused by retrograde pressure with 100 ml cardioplegia solution (lactated Ringer's supplemented with 20 m E q / L potassium chloride) via the cross-clamped descending aorta. This procedure was followed by pressure perfusion with 100 cc 10 % buffered formalin. The heart and lungs were then placed in formalin solution. The heart was dissected free of all attached pulmonary and vascular tissue and epicardial fat. The distal terminus of the pericardial catheter was palpated and marked on the pericardium with india ink. The catheter was then removed, and the heart was cut in four transverse (shortaxis) sections from apex to base. The septal and posterior wall thicknesses were measured at the midpapillary muscle level and were recorded. Each section was submitted for staining with hematoxylin-eosin and with the aniline blue trichrome stain for collagen. Histologic analysis. Microvascular density and formation of new capillaries were assessed by histopathologic examination of subepicardial myocardial tissue performed by an examiner in a blinded fashion and were graded as fol10ws: (0) normal vascularity, minimal or no epicardial inflammatory response; (1) mild to moderate inflammatory response, some new vessel formation; and (2) marked increase in vascularity with communication of vascular channels from pericardium to myocardium. The presence of myocardial necrosis in hearts with hypertrophy is a cot-
Table Ih Left ventricular wall thickness Intravenous AII (groups 1,3,4, n=19)
Septum (mm) Posterior wall (ram) Septum + posterior wall (ram) Septum/posterior wall ratio
Normal saline solution (group 2, p Value n=5)
5.62 ± 1.00 5.42 _+ 1.12
3.98 _+ 0.61 4.68 ± 0.78
<0.005 <0.15
11.04± 2.00
8.67 ± 1.32
<0.02
1.05 _+ 0.13
0.86 _+ 0.10
<0.005
relate for the degree of chronic ischemia experienced in these preparations. 2° Each of the four sections from each heart was examined in its entirety under low-power (x40) magnification to assess the severity of necrosis present. This phenomenon was quantified by an examiner in a blinded fashion on a 0 to 3 scale in trichrome-stained specimens as follows: (0) no fibrotic response, collagen in perivascular locations only; (1) minimal increase in interstitial fibrosis, manifest in less t h a n 50 % of the low-power fields examined; (2) moderate interstitial fibrosis, present in more t h a n half of each section examined; and (3) severe interstitial fibrosis, present in virtually all fields with coalescent islands of scar replacing large segments of myocardium. Statistics. The four groups were compared by the Kruskall-Wallis statistic for ordinal variables. For continuous variables a one-way analysis of variance was performed with pairwise post hoc analysis by Scheffe's method to determine significant differences among groups. For ordinal variables pairwise analysis was performed with the M a n n - W h i t n e y U test. Statistical analysis was carried out with the Statview II (version 1.03) software package (Abacus Concepts, Berkeley, Calif.) r u n n i n g on a Macintosh IIci (Apple Computer, Inc., Cupertino, Calif.) computer. All data are reported as mean z SD.
RESULTS Characteristics of the experimental groups. A n i m a l s i n each of t h e four g r o u p s r e c e i v e d b e t w e e n 7 a n d 28 d a y s of i n t r a p e r i c a r d i a l or i n t r a v e n o u s i n f u s i o n s . T h e
May 1 9 9 5
928
Landau, Jacobs, and Haudenschild
American Heart Journal
E
.J ,,.,.
x'x x'x ~X x'x ~X xx xx xx xx xx x'x ~x
"5 ,.O
E
x~x x~x ~X
U
I
Group 1
012 Group 2
012 Group3
012 Group 4
Angiogenesis score' Fig. 3. Distribution ofneovascular response scores in four treatment groups.
Fig. 2. Representative samples of neovascularization response scores of 0 through 2. In all sections epicardial surface (E) is oriented superiorly. A, Score of 0 was assigned to this section with normal epicardial vascularity. B, Augmented vascularity is present with increased numbers of vascular channels in the epicardial space, yielding a score of 1. C, Score of 2 was given because of a marked increase in vascularity with large vascular channels in the periepicardial region and direct communication between the myocardium and surrounding pericardial vascular structures (arrow). (Hematoxylin-eosin stain, original magnification ×40.)
number of infusion days, dose of angiotensin II used, and maximum measured systolic blood pressure are recorded in Table I. The dose of AII used is in the range that induces left ventricular hypertrophy in the absence of hemodynamic effects.24, 25 No significant differences exist among the experimental groups in these parameters. Characteristics of all-induced ventricular hypertrophy. The ability of intravenous angiotensin II infusions to promote rapid ventricular hypertrophy and resultant ischemia was manifest in comparisons of hearts from groups 1, 3, and 4 (AII infusions) versus hearts in group 2 (normal saline solution infusion). Septal and posterior wall thicknesses and their sum and ratio are shown in Table II. Animals in the angiotensin groups consistently displayed left ventricular hypertrophy that disproportionately affected the interventricular septum (Septal/Posterior wall thickness ratio >1) compared with animals in the saline solution control group. No medial hypertrophy of arterioles was observed in sections from any of the four groups, as was described in canine experiments that used epicardial aFGF. 19 Histologic evidence of inadequate vascular supply in hearts with hypertrophy was reflected in the degree of myocyte replacement by scar tissue. Mean fibrosis scores of rabbits treated with angiotensin II was significantly greater than those of rabbits receiving intravenous saline solution (1.65 _+ 0.96 vs 0.22 _+ 0.40, p < 0.005), supporting the hypothesis
Volume 129, Number 5
Landau, Jacobs, and Haudenschild
AmericanHeartJournal
929
Table III. Angiogenesis scores No. of animals with score = 2 Group
1 2 3 4
Intervention
IP IP IP IV
b F G F + IV AII b F G F + IV saline solution Alb + IV AII AII only
No.
7 5 7 5
(%)
M e a n score*
p Values vs group 4
p Values vs group 3
6 (86) 2 (40) 3 (43)
1.9 1.4 1.3
<0.003 <0.02 <0.03
<0.10 <0.90
0 (0)
o.2
Abbreviations as in Table I. *p < 0.001 by analysis of variance among all groups
that AII-induced hypertrophy resulted in microinfarcts caused by severe myocardial oxygen supply/ demand mismatch. Angiogenic responses to intrapericardial infusions.
Fig. 2 shows representative sections from animals with vascularity scores of 0, 1, and 2, respectively. The angiogenic response was most marked in the region of the distal termination of the intrapericardial catheter, and therefore the angiogenesis score assigned to an animal was the maximum of the four sections examined for each heart. In group 4 (without an intrapericardial catheter) the same protocol was followed. The distribution of angiogenesis scores is displayed in Fig. 3. Six of seven rabbits receiving intrapericardial bFGF and intravenous AII (group 1) demonstrated a marked increase in vascularity, and intermediate degrees of angiogenesis were seen in the two other sets of animals treated with intrapericardial infusions (groups 2 and 3). Animals receiving intravenous AII only (group 4) had normal epicardial vascutarity, with one of five hearts demonstrating minimal augmentation in vascular density. The mean angiogenesis scores and percentage of animals with a marked angiogenic response (angiogenesis score = 2) for the four groups are shown in Table III. A trend towards increased vascularity was observed in group 1 versus group 3 (angiogenesis scores of 1.9 and 1.3, respectively, p < 0.10), suggesting that prolonged bFGF exposure augments new epicardial vessel growth in this hypertrophy model. DISCUSSION Study results. These experiments demonstrate ev-
idence of new capillary formation in both groups of animals receiving intrapericardial bFGF. A more pronounced effect was observed in animals receiving concomitant angiotensin II, suggesting a potentiating effect from myocardial ischemia associated with the severe left ventricular hypertrophy observed in
this group of animals. The replacement fibrosis that was prominently seen in animals treated with AII may also have been a result of a direct toxic effect of angiotensin, 25 leading to microinfarction and scarring. The increased angiogenesis score of group 3 compared with that of group 4 is likely due to an inflammatory response to the intrapericardial infusion tube or the bovine albumin infusate. Nonetheless in the presence of AII, bFGF augmented the angiogenic response compared with albumin (group I v s group 3). These results are consistent with the known anglogenic activity of bFGF in a variety of tissues and with more recent observations that intracoronary bFGF induces new vessel formation in ischemic myocardial tissue.5, is The lack of enhanced epicardial vascularity in hearts in group 4 is in agreement with several reports documenting lack of capillary growth associated with the development of myocardial hypertrop h y . 20-22
bFGF is one of a family of growth factors known to have a high affinity for heparin. 26 The highly localized effect of bFGF in promoting angiogenesis may be due to the avid binding of this peptide to basement membrane and extracellular matrix in the region of the infusion catheter terminus caused by the high content of heparin-like molecules in these tissues. 27 Comparison with other investigations. The angiogenic potential of intracoronary bFGF has been assessed by several investigators. In a canine infarct model Yanagisawa-Miwa et al. 5 have shown that a single intracoronary infusion of bFGF increased the number of arterioles and capillaries in the infarct border zone after 1 week. The authors observed an apparent reduction in infarct size in animals treated with bFGF and proposed that this reduction resulted from bFGF-induced formation of new microvasculature. In a similar model Unger et al. 2s demonstrated enhanced collateral flow in dogs exposed to daily in-
May 1995
930
Landau, Jacobs, and Haudenschild
tracoronary bFGF for 28 days, also supporting an angiogenic effect of this peptide in myocardium. In a porcine model of microinfarctions induced by intracoronary injection of agarose beads, animals were exposed to bFGF, which was adsorbed to the beads and was presumably released locally. Higher microvessel counts were observed in animals treated with bFGF compared with animals in the control group 2 weeks after infarction. The difference was most marked in sections with extensive evidence of nonviable myocardium, ls As in this study the presence of myocardial necrosis appears to have enhanced the degree of angiogenesis produced by bFGF. Banal et al. 19 applied high doses of sponge-embedded acidic FGF (aFGF) topically to the epicardium in both ischemic and nonischemic dogs and assessed collateral development from an internal mammary artery pedicle placed over the sponge. No new vessel growth was observed, although pronounced smoothmuscle cell proliferation within the wall of arterioles and small arteries in the infarcted territories of animals treated with aFGF was noted. 19 We observed no such phenomenon possibly because of the use of a different growth factor, lower doses, species-related differences, or the fact that tissue fibrosis in this study was patchy. The angiogenic effects of bFGF have also been evaluated in skeletal muscle beds. Chleboun et al. s implanted ethylene-vinyl acetate pellets that were impregnated with bFGF into rat hind limbs rendered ischemic by ligation of the common iliac, common femoral, and superficial femoral arteries. The impregnated pellets released bFGF in vitro for a period of at least 24 hours. After 3 weeks rats treated with bFGF had improved arterial flow measured by a Doppler technique compared with the control group, although the difference became nonsignificant at the 4-week observation point. The authors did not perform any direct analysis of the degree of collateral flow, nor was a histologic analysis undertaken to characterize the mechanism of the improved arterial flow in the bFGF group. Baffour 9 used a similar model in the rabbit, administering i or 3 ttg bFGF via daily intramuscular injections at four sites in the thigh and one site in the calf for a 2-week period. Response to therapy was assessed with transcutaneous oximetry, angiography, and clinical limb status. The authors found a doserelated improvement in these indexes of muscle perfusion and viability in the bFGF group compared with those of the control group. Similar results have also been demonstrated with intramuscular injec-
AmericanHeartJournal
tions of endothelial cell growth factor, an acidic FGF, and heparin. 29 In conclusion, exposure of a variety of tissues to locally delivered bFGF results in new capillary formation in vivo. Moreover the extent of bFGF-induced angiogenesis was sufficient to alter indexes of tissue performance to a physiologically significant degree in some models of vascular insufficiency. 5, 8, 9 These observations suggest that local exposure of muscle beds to bFGF could produce physiologically important effects on tissue perfusion and suggest the capability for salutary effects in ischemic tissues. Conclusions. We have demonstrated the ability of locally administered bFGF to induce new vessel growth when applied within the pericardial space. This series of experiments also suggests that ischemia facilitates bFGF-induced myocardial angiogenesis in this model. Whether the enhanced vascular supply can be sustained after bFGF is no longer provided is an important issue that will require additional study. Study limitations. Although neovascularization was observed in this model, the source of new vessel growth could not be definitively identified with the methods used. The left ventricular hypertrophy model used is presumed to be ischemic; however, the replacement fibrosis observed may be the result of direct angiotensin II myocyte toxicity rather than infarction caused by supply/demand mismatch. This finding does not alter the study findings that in the setting of inadequate capillary supply caused by left ventricular hypertrophy, bFGF efficiently promotes new vessel growth. REFERENCES
1. Canty JM, Klocke FJ. Reduced regional myocardial perfusion in the presence of pharmacologic vasodilator reserve. Circulation 1985;71:370-7. 2. Landau C, Jacobs AK, Haudenschild CC. Left ventricular hypertrophy induced by angiotensin II is accompanied by a dose dependent fibrotic response [Abstract]. Circulation 1992;86[suppl I]:I-754. 3. Yancy CW, Vissing S, Cuckey JC, Bellomo JF, Firth BG, Blomqvist CG. Maximal conductance versus maximal oxygen uptake in patients with congestive heart failure [abstract]. J Am Coll Cardiol 1988;ll[suppl A]:73A. 4. Halperin JL, creager MA. Arterial obstructive diseases of the extremeties. In: Loscalzo J, Creager MA, Dzau VJ, eds. Vascular medicine. Boston: Little, Brown and Company; 1992:835-43. 5. Yanagisawa-Miwa A, Uchida Y, Nakamura F, Tomaru T, Kido H, Kamijo T, Sugimoto T, Kaji K, Utsuyama M, Kurashima C, Ito H. Salvage of infarcted myocardium by angiogenic action of basic fibroblast growth factor. Science 1992;257:1401-3. 6. Chilian WM, Ackell PA. Nonuniform transmural alpha-adrenergic coronary constriction in the presence of a stenosis. Circ Res 1981;48:41623. 7. Martin WH, Montgomery J, Snell PG, Sokolov JJ, Buckey JC, Maloney DA, Blomqvist CG. Cardiovascular adaptations to intense swim training in sedentary middle-aged humans. Circulation 1987;75:323-30.
Volume 129, Number 5 American Heart Journal
8. Chleboun JO, Martins RN, Mitchell CA, Chirfla TV. bFGF enhances the development of collateral circulation after acute arterial occlusion. Biochem Biophys Res Commun 1992;185:510-6. 9. Baffour R, Berman J, Garb JL, Rhee SW, Kaufman J, Friedmann P. Enhanced angiogenesis and growth of collaterals by in vivo administration of recombinant basic fibroblast growth factor in a rabbit model of acute lower limb ischemia: dose-response effect of basic fibroblast growth factor. J Vasc Surg 1992;16:181-91. 10. Gospodarowicz D, Ferrara N, Schweigerer L, Neufeld G. Structural characterization and biological functions of fibroblast growth factor. Endocr Rev 1987;8:95-114. 11. Muller R, Bravo R, Burkshardt J, Curran T. Induction of c los gene and protein by growth factors precedes activation of c myc. Nature 1984;312:716-9. 12. Cuevas P, Gonzalez AM, Carceller F, Baird A. Vascular response to basic fibroblast growth factor when infused into the normal adventitia or into the injured media of the rat carotid artery. Circ Res 1991;69:360-9. 13. Klagsbrun M, Edalman ER. Biological and biochemical properties of fibroblast growth factors: implications for the pathogenesis of atherosclerosis. Arteriosclerosis 1989;9:269-78. 14. Thompson JA, Anderson KD, DiPietro JM, Zwiebel JA, Zametta M, Anderson WF, Maciag T. Site-directed neovessel formation in vivo. Science 1988;241:1349-52. 15. Thompson JA, Handenschild CC, Anderson KD, DiPietro JM, Anderson WF, Maciag T. Heparin binding growth factor 1 induces the formation of organoid neovascular structures in vivo. Proc Natl Acad Sci U S A 1989;86:7928-32. 16. Eppley BL, Doucet M, Connolly DT, Eerier J. Enhancement of anglogenesis by basic fibroblast growth factor in mandibular bone graft healing in the rabbit. J Oral Maxillofac Surg 1988;46:391-8. 17. Hayek A, Culler FL, Beattie GM, Lopez AD, Cuevas P, Baird A. An in vivo model for the study of the angiogenic effects of basic fibrobIast growth factor. Biochem Biophys Res Commun 1987;147:876-80. 18. Battier A, Scheinowitz M, Bor A, Hasdai D, Vered Z, Di Segni E, Varda-BIoom N, Nass D, Engelberg S, Eldar M, Belkin M, Savion N. Intracoronary injection of basic fibroblast growth factor enhances angio-
Landau, Jacobs, and Haudenschild
19.
20.
21.
22. 23.
24.
25. 26. 27.
28.
29.
931
genesis in infarcted swine myocardium. J Am Coll Cardiol 1993;22: 2001-6. Banal S, Jaklitsch MT, Casscells W, Shou M, Shrivasfar S, Correa R, Epstein SE, Unger EF. Effects of acidic fibroblast growth factor on norreal and ischemie myocardium. Circ Res 1991;69:76-85. Rakusan K, Flanagan MD, Geva T, Southern J, Van Praagh R. Morphometry or human coronary capillaries during normal growth and the effect of age in left ventricular pressure-overload hypertrophy. Circulation 1992;86:38-46. Breisch EA, White FC, Nimmo LE, Bloor CM. Cardiac vasculature and flow during pressure-overload hypertrophy. Am J Physiol 1986; 251:H1031-7. Hudlicka O. Growth of capillaries in skeletal and cardiac muscle. Circ Res 1982;50:451-6L Edelman ER, Mathiowitz E, Langer R, Klagsbrun M. Controlled and modulated release of basic fibroblast growth factor. Biomaterials 1991;12:619-26. Brown J J, Chapuis G, Robertson JIS. The effect of prolonged intravenous infusion of angiotensin in the rabbit. Clin Sci (Colch) 1964;26:16575. Tan L, Jalil JE, Rick R, Janicki JS, Weber KT. Cardiac myocyte necrosis induced by angiotensin II. Circulation 1992;69:1185-95. Burgess WH, Maciag T. The heparin-binding (fibroblast) growth factor family of proteins. Annu Rev Biochem 1989;58:575-606. Vlodavsky I, Folkman J, Sullivan R, Fridman R, Ishai-Michaeli R, Sasse J, Klagsbrun M. Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix. Proc Natl Acad Sci U S A 1987;84:2292-6. 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 blood flow in a canine model. Am J Physiol 1994;266:H1588-95. Pu L-Q, Sniderman AD, Brassard R, Lachapelle KJ, Graham AM, Lisbona R, Symes JF. Enhanced revascularization of the ischemic limb by angiogenic therapy. Circulation 1993;88:208-15.
Charles Landau, MD, Alice K. Jacobs, MD, and Christian C. Haudenschild, MD Boston, Mass.
Compromised microvascular circulation is an important component of several cardiovascular diseases including syndromes of acute coronary ischemia, 1 hypertensive heart disease with left ventricular hypertrophy, 2 chronic congestive heart failure, 3 and diabetic peripheral vascular insufficiency. 4 In each of these conditions an unfavorable or inappropriate ratio of capillary density to tissue oxygen demand appears to contribute to the extent, severity, or persistence of muscle dysfunction. 36 In theory interventions to increase capillary density in underperfused tissue beds could have a meaningful palliative effect Prom the Evans Memorial Department of Clinical Research and the Section of Cardiology, Department of Medicine, Boston University Medical Center (CL, AKJ), and the Mallory Institute of Pathology (CCH), Boston. Supported by a Biomedical General Research Support Grant of University Hospital, Boston. Received for publication Aug. 25, 1994; accepted Oct. 7, 1994. Reprint requests: Charles Landau, MD, University of Texas Southwestern Medical Center, Cardiology Division, 5323 Harry Hines Blvd., Dallas, TX 75235-9047. Copyright © 1995 by Mosby-Year Book, Inc. 0002-8703/95/$3.00 + 0
924
4/1/61996
in these conditions. This possibility is supported by several observations including (1) a correlation between improvement in functional class and skeletal muscle capillary density in patients with chronic congestive heart failure subjected to graded exercise programs, 7 (2) an improvement in postinfarction left ventricular systolic function in experimental animals as a result of myocardial angiogenesis induced by administration of exogenous basic fibroblast growth factor (bFGF), an angiogenic peptide, 5 and (3) decreased skeletal muscle ischemia in animals receiving local administration of bFGF in a rabbit hind limb model of vascular insufficiency. 8, 9 Basic fibroblast growth factor (bFGF) is a 17 kd cationic polypeptide consisting of 154 amino acids. Basic FGF is mitogenic for a number of primary and established cell types in culture including vascular endothelial cells. 1° Although the molecular mechanism by which bFGF exerts its mitogenic effect is incompletely defined, cells exposed to bFGF demonstrate increased expression of several cellular protooncogenes including c-myc. 11 Basic FGF also has
Volume 129, Number 5 American Heart Journal
Landau, Jacobs, and Haudenschild 925 New Zealand White Rabbits
Implan
.
.
.
Implantation of angiotensin 1I (All) osmotic pump into external jugular vein
Intrapericard aI(IP)bFGF
1-4 w e e k s
Grou ~1 IP bFGF+IV All n=7
.
Implantation of saline osmotic p u m p into external jugular vein
Intraperic~ rdial(IP)bFGF
Intraperk rdial(IP) Albumin
1-4 weeks
1-4 weeks
Grou ~2 IP bFGF+W Saline n=5
Implantation of angiotensin I1 (All) osmotic p u m p into external jugular vein
Grou, 3 IP Alb + IV AlI n=7
1-4 weeks
Gro~, 4 1V AII only n=5
Sacrifice,Neovascular response score
Fig. 1. Flow diagram of animal protocol. IP, Intrapericardial; IV, intravenous; AII, angiotensin II; bFGF, basic fibroblast growth factor; Alb, albumin.
pleiotropic effects on the promotion of capillary growth and invasion. 1°13 It has been shown to induce angiogenesis in several animal models. Applied directly or adsorbed to implantable synthetic materials that release the peptide over time, bFGF has been demonstrated to induce angiogenesis in chick chorioallantoic membrane, ~° rabbit cornea, ~° rat peritoneal cavity, 14,15 rat neck, 14 rabbit mandible, 16 rat perinephric capsule, 17 and in the adventitia or media of rat carotid arteries. 12 In each of these experiments prolonged exposure to bFGF (at least 1 week) appeared to be required for an angiogenic response. Neovascularization has also been described in skeletal and cardiac muscle exposed to bFGF in the setting of ischemia. 5, 8, 9,18 In theory locally applied angiogenic factors could potentially enhance the efficacy of surgical or percutaneous revascularization procedures by promoting new vessel growth in ischemic regions. In a canine model of myocardial ischemia and infarction, acidic fibroblast growth factor (aFGF) applied to the epi-
cardial surface did not demonstrate evidence of new capillary growth, 19 but the effects o f bFGF in this setting have not been evaluated previously. Accordingly this study was undertaken to assess the ability of bFGF to induce angiogenesis when applied to the pericardial space under conditions of increased myocardial oxygen demand. METHODS
All procedures were performed in male New Zealand White rabbits under the guidelines of the Animal Welfare Act. The protocol was approved by the Institutional Animal Care and Use Committee of the Boston University Medical Center. Animals received continuous infusion osmotic pumps that delivered intravenous solutions, intrapericardial solutions, or both for periods of up to 28 days. It is well established that acquired left ventricular hypertrophy in adult animals is associated with inadequate compensatory capillary growth. 2°,21Intravenous angiotensin II (AII) was used to rapidly induce left ventricular hypertrophy,2 which provided an ischemic milieu caused by the inadequate density of capillaries present to supply the
May 1995
926
Landau, Jacobs, and Haudenschild
additional myocardium. 22 Rabbits were assigned to one of four groups: group 1 received intrapericardial bFGF and intravenous AII, group 2 received intrapericardial bFGF and intravenous saline solution, group 3 received intrapericardial albumin and intravenous AII, and group 4 received intravenous AII only. Tail cuff blood pressures were measured twice weekly for the duration of the experiment. A flow diagram (Fig. 1) summarizes the protocol sequence. Intrapericardial catheter placement, New Zealand White rabbits weighing 2.0 to 4.3 kg were quarantined for a period of 1 week to exclude communicable infections. Because pilot studies revealed a 10 % mortality rate after pericardial instrumentation, the initial procedure in groups 1 through 3 was implantation of a pericardial catheter to create a portal for intrapericardial infusions. To achieve this goal animals were anesthetized with xylazine (5 mg/kg) and ketamine (35 mg/kg) given subcutaneously. The subxiphoid region was shaved and prepped in sterile fashion. The animal was placed in the supine position and was rotated slightly to the left so that the heart rested to the left of midline under fluoroscopy. Lidocaine (Xylocaine) (2 %) was infused subcutaneously, and a longitudinal 3 cm incision was made in the midline with the superior aspect overlying the xiphoid process. Blunt dissection was undertaken, forming a tunnel through superficial layers. A 22-gauge spinal needle placed within a 5F dilator was inserted through the tunnel. The proximal end of the needle was connected to a tuohy adapter through which a 0.014-inch diameter Flex guide wire (USCI, Billerica, Mass.) was placed. The guide wire was positioned so that the distal end was 1 cm proximal to the needle tip. A 10 ml syringe filled with 70 % diatrizoate Renografin-76, Squibb, Princeton, N.J.) was attached to the infusion port of the tuohy adapter. The needle assembly was then directed at a 15-degree angle to the skin surface and was advanced under fluoroscopy towards the cardiac silhouette to the left of the midline. It was advanced gently until it popped through the diaphragm into the pericardial space. Correct needle position was confirmed by pulsatile motion of the needle tip. Contrast was injected to assess the location of the distal needle. An intraventricular location was indicated by rapid dye clearance, whereas an intrapericardial location of the needle was indicated by gravity-dependent pooling of contrast in the pericardial space. Dye staining represented an intramyocardial or intrapleural location. If necessary, the needle was repositioned with repeated contrast injections until a pericardial location was achieved. The guide wire was advanced into the pericardial space as confirmed by the wire remaining within the cardiac silhouette. The 5F dilator was passed over the wire, and the needle and wire were removed. The location of the dilator was confirmed with a contrast injection. Next a 0.018-inch diameter Flex guide wire (USCI) was passed through the dilator and was looped superiorly and then inferiorly around the heart so that the distal end rested in the region of the anterolateral wall of the left ventricle. The dilator was removed, and a 20 cm length of polyethylene tubing (PE-60, Fisher Scientific, Santa Clara,
AmericanHeartJournal
Calif.) was advanced over the wire to the distal end of the wire. After the wire was removed, position was confirmed by contrast injection through the tubing. Saline solution was flushed through the catheter, and it was anchored to abdominal muscle fascia with 4-0 silk suture. The proximal end of the catheter was placed in a subcutaneous pocket, and the skin was closed with absorbable suture. Animals received 100,000 U intramuscular penicillin on the day of the procedure and on alternate days for a total of three doses. Neck pump placement. Two days after intrapericardial catheter placement was done, osmotic pumps (Alza Model 2ML4, Alza Corp, Palo Alto, Calif.) were loaded in sterile fashion with the appropriate dose of angiotensin II (AII) or normal saline solution (Fig. 1) and were attached to a 5 cm length of PE-60 tubing. The pumps were incubated at 37 ° C for a minimum of 4 hours to activate infusion activity as specified by the manufacturer. Animals were anesthetized as described previously. The neck region was shaved and draped in sterile fashion with 2 % Xylocaine instilled subcutaneously. A midline 3 to 4 cm neck incision was made, the external jugular vein was isolated, and a subcutaneous pocket was formed. The distal aspect of the vein was tied off, and a venotomy was made. The pump was removed from the heated bath and positioned in the pocket. The attached tubing was trimmed so that approximately 2 cm of catheter could be advanced into the vein. An intravenous dose of 200 U heparin was given, the tubing was secured into its intravenous location with silk ties, and the pocket and skin were sutured closed. Antibiotics were administered as indicated previously. Because of the dehydrating effects of the AII, all animals with AII pumps were weighed three times per week and received 5 % dextrose in 0.9 % normal saline solution at a rate of 30 ml/day by subcutaneous infusion. If weight loss was evident, this administration was increased to 50 ml/day, until animal weight was stabilized. After I week of intravenous AII or saline solution infusion, the pericardial pump was attached in animals in groups 1 through 3 to the previously placed catheter as described in the following section. Pericardial pump attachment. Pericardial pumps (Alza, model 2ML4) were loaded in sterile fashion with 5/~g basic human recombinant FGF (Gibco, Grand Island, N.Y.) diluted to a volume of 2 ml with 0.1% bovine serum albumin (BSA) in phosphate-buffered saline solution filtered through a 0.45 ~m millipore filter. The resultant dose of bFGF was 180 ng/day. It has been demonstrated that bFGF retains its biologic activity at 37 ° C for periods of up to 4 weeks when dispensed from a sequestered source. 23Animal pumps in the control group were loaded with 0.1% BSA only (Fig. 1). The rabbit was anesthetized as described previously. The previously placed subxiphoid sutures were removed, and the wound was partially opened. The pericardial catheter was identified and isolated, and its position in the pericardial space was reconfirmed with a contrast injection. A subcutaneous pocket was created with blunt dissection.
Volume 129, Number 5 AmericanHeartJournal
927
Landau, Jacobs, and Haudenschild
Table I. Baseline characteristics
AII dose (~g/kg/hr) Days of AII infusion Days of IP infusion Blood pressure (mm Hg)
Group 1 (IP bFGF + I V AII, n = 7)
Group 2 (IP bFGF + I V saline solution, n = 5)
Group 3 (IP Alb + I V AII, n = 7)
Group 4 (IV AII, n = 5)
p Value
17.8 + 6.8 24.3 _+ 6.8 22.7 + 9.2 130 ± 20
NA NA 28 _+ 0 113 __ 12
18.7 + 8.6 20.9 _+ 7.9 18 + 9.9 131 ± 16
19.8 ± 12.1 16.6 ± 7.3 NA 109 _+ 26
NS NS NS NS
IP, Intrapericardial; IV, intravenous; Alb, albumin; A/I, angiotensin II; bFGF, basic fibroblast growth factor; NS, not significant; NA, not applicable.
The p u m p was placed in the pocket and was attached to the pericardial tubing. The tubing was anchored to the pump with silk ties, and the pump was then sutured to subcutaneous fascia. The skin was closed, and antibiotics were administered as previously described. Two weeks after the p u m p was implanted, the animal was again anesthetized, and the pump was exposed to confirm that it was still adequately positioned in the pocket and was connected to the pericardial tubing. Euthanization. Four weeks after pericardial pump placem e n t the animal was again anesthetized. A midline incision from the sternal notch to the abdominal region adjacent to the pericardial pump was made. Pentothal was given in a lethal intravenous dose of 25 mg/kg. The heart was exposed by a median sternotomy, and the pericardial catheter was carefully located. Its proximal portion was ligated in the supradiaphragmatic region so that its terminal position in the pericardial space was preserved. The heart and lungs were quickly removed, and the heart was perfused by retrograde pressure with 100 ml cardioplegia solution (lactated Ringer's supplemented with 20 m E q / L potassium chloride) via the cross-clamped descending aorta. This procedure was followed by pressure perfusion with 100 cc 10 % buffered formalin. The heart and lungs were then placed in formalin solution. The heart was dissected free of all attached pulmonary and vascular tissue and epicardial fat. The distal terminus of the pericardial catheter was palpated and marked on the pericardium with india ink. The catheter was then removed, and the heart was cut in four transverse (shortaxis) sections from apex to base. The septal and posterior wall thicknesses were measured at the midpapillary muscle level and were recorded. Each section was submitted for staining with hematoxylin-eosin and with the aniline blue trichrome stain for collagen. Histologic analysis. Microvascular density and formation of new capillaries were assessed by histopathologic examination of subepicardial myocardial tissue performed by an examiner in a blinded fashion and were graded as fol10ws: (0) normal vascularity, minimal or no epicardial inflammatory response; (1) mild to moderate inflammatory response, some new vessel formation; and (2) marked increase in vascularity with communication of vascular channels from pericardium to myocardium. The presence of myocardial necrosis in hearts with hypertrophy is a cot-
Table Ih Left ventricular wall thickness Intravenous AII (groups 1,3,4, n=19)
Septum (mm) Posterior wall (ram) Septum + posterior wall (ram) Septum/posterior wall ratio
Normal saline solution (group 2, p Value n=5)
5.62 ± 1.00 5.42 _+ 1.12
3.98 _+ 0.61 4.68 ± 0.78
<0.005 <0.15
11.04± 2.00
8.67 ± 1.32
<0.02
1.05 _+ 0.13
0.86 _+ 0.10
<0.005
relate for the degree of chronic ischemia experienced in these preparations. 2° Each of the four sections from each heart was examined in its entirety under low-power (x40) magnification to assess the severity of necrosis present. This phenomenon was quantified by an examiner in a blinded fashion on a 0 to 3 scale in trichrome-stained specimens as follows: (0) no fibrotic response, collagen in perivascular locations only; (1) minimal increase in interstitial fibrosis, manifest in less t h a n 50 % of the low-power fields examined; (2) moderate interstitial fibrosis, present in more t h a n half of each section examined; and (3) severe interstitial fibrosis, present in virtually all fields with coalescent islands of scar replacing large segments of myocardium. Statistics. The four groups were compared by the Kruskall-Wallis statistic for ordinal variables. For continuous variables a one-way analysis of variance was performed with pairwise post hoc analysis by Scheffe's method to determine significant differences among groups. For ordinal variables pairwise analysis was performed with the M a n n - W h i t n e y U test. Statistical analysis was carried out with the Statview II (version 1.03) software package (Abacus Concepts, Berkeley, Calif.) r u n n i n g on a Macintosh IIci (Apple Computer, Inc., Cupertino, Calif.) computer. All data are reported as mean z SD.
RESULTS Characteristics of the experimental groups. A n i m a l s i n each of t h e four g r o u p s r e c e i v e d b e t w e e n 7 a n d 28 d a y s of i n t r a p e r i c a r d i a l or i n t r a v e n o u s i n f u s i o n s . T h e
May 1 9 9 5
928
Landau, Jacobs, and Haudenschild
American Heart Journal
E
.J ,,.,.
x'x x'x ~X x'x ~X xx xx xx xx xx x'x ~x
"5 ,.O
E
x~x x~x ~X
U
I
Group 1
012 Group 2
012 Group3
012 Group 4
Angiogenesis score' Fig. 3. Distribution ofneovascular response scores in four treatment groups.
Fig. 2. Representative samples of neovascularization response scores of 0 through 2. In all sections epicardial surface (E) is oriented superiorly. A, Score of 0 was assigned to this section with normal epicardial vascularity. B, Augmented vascularity is present with increased numbers of vascular channels in the epicardial space, yielding a score of 1. C, Score of 2 was given because of a marked increase in vascularity with large vascular channels in the periepicardial region and direct communication between the myocardium and surrounding pericardial vascular structures (arrow). (Hematoxylin-eosin stain, original magnification ×40.)
number of infusion days, dose of angiotensin II used, and maximum measured systolic blood pressure are recorded in Table I. The dose of AII used is in the range that induces left ventricular hypertrophy in the absence of hemodynamic effects.24, 25 No significant differences exist among the experimental groups in these parameters. Characteristics of all-induced ventricular hypertrophy. The ability of intravenous angiotensin II infusions to promote rapid ventricular hypertrophy and resultant ischemia was manifest in comparisons of hearts from groups 1, 3, and 4 (AII infusions) versus hearts in group 2 (normal saline solution infusion). Septal and posterior wall thicknesses and their sum and ratio are shown in Table II. Animals in the angiotensin groups consistently displayed left ventricular hypertrophy that disproportionately affected the interventricular septum (Septal/Posterior wall thickness ratio >1) compared with animals in the saline solution control group. No medial hypertrophy of arterioles was observed in sections from any of the four groups, as was described in canine experiments that used epicardial aFGF. 19 Histologic evidence of inadequate vascular supply in hearts with hypertrophy was reflected in the degree of myocyte replacement by scar tissue. Mean fibrosis scores of rabbits treated with angiotensin II was significantly greater than those of rabbits receiving intravenous saline solution (1.65 _+ 0.96 vs 0.22 _+ 0.40, p < 0.005), supporting the hypothesis
Volume 129, Number 5
Landau, Jacobs, and Haudenschild
AmericanHeartJournal
929
Table III. Angiogenesis scores No. of animals with score = 2 Group
1 2 3 4
Intervention
IP IP IP IV
b F G F + IV AII b F G F + IV saline solution Alb + IV AII AII only
No.
7 5 7 5
(%)
M e a n score*
p Values vs group 4
p Values vs group 3
6 (86) 2 (40) 3 (43)
1.9 1.4 1.3
<0.003 <0.02 <0.03
<0.10 <0.90
0 (0)
o.2
Abbreviations as in Table I. *p < 0.001 by analysis of variance among all groups
that AII-induced hypertrophy resulted in microinfarcts caused by severe myocardial oxygen supply/ demand mismatch. Angiogenic responses to intrapericardial infusions.
Fig. 2 shows representative sections from animals with vascularity scores of 0, 1, and 2, respectively. The angiogenic response was most marked in the region of the distal termination of the intrapericardial catheter, and therefore the angiogenesis score assigned to an animal was the maximum of the four sections examined for each heart. In group 4 (without an intrapericardial catheter) the same protocol was followed. The distribution of angiogenesis scores is displayed in Fig. 3. Six of seven rabbits receiving intrapericardial bFGF and intravenous AII (group 1) demonstrated a marked increase in vascularity, and intermediate degrees of angiogenesis were seen in the two other sets of animals treated with intrapericardial infusions (groups 2 and 3). Animals receiving intravenous AII only (group 4) had normal epicardial vascutarity, with one of five hearts demonstrating minimal augmentation in vascular density. The mean angiogenesis scores and percentage of animals with a marked angiogenic response (angiogenesis score = 2) for the four groups are shown in Table III. A trend towards increased vascularity was observed in group 1 versus group 3 (angiogenesis scores of 1.9 and 1.3, respectively, p < 0.10), suggesting that prolonged bFGF exposure augments new epicardial vessel growth in this hypertrophy model. DISCUSSION Study results. These experiments demonstrate ev-
idence of new capillary formation in both groups of animals receiving intrapericardial bFGF. A more pronounced effect was observed in animals receiving concomitant angiotensin II, suggesting a potentiating effect from myocardial ischemia associated with the severe left ventricular hypertrophy observed in
this group of animals. The replacement fibrosis that was prominently seen in animals treated with AII may also have been a result of a direct toxic effect of angiotensin, 25 leading to microinfarction and scarring. The increased angiogenesis score of group 3 compared with that of group 4 is likely due to an inflammatory response to the intrapericardial infusion tube or the bovine albumin infusate. Nonetheless in the presence of AII, bFGF augmented the angiogenic response compared with albumin (group I v s group 3). These results are consistent with the known anglogenic activity of bFGF in a variety of tissues and with more recent observations that intracoronary bFGF induces new vessel formation in ischemic myocardial tissue.5, is The lack of enhanced epicardial vascularity in hearts in group 4 is in agreement with several reports documenting lack of capillary growth associated with the development of myocardial hypertrop h y . 20-22
bFGF is one of a family of growth factors known to have a high affinity for heparin. 26 The highly localized effect of bFGF in promoting angiogenesis may be due to the avid binding of this peptide to basement membrane and extracellular matrix in the region of the infusion catheter terminus caused by the high content of heparin-like molecules in these tissues. 27 Comparison with other investigations. The angiogenic potential of intracoronary bFGF has been assessed by several investigators. In a canine infarct model Yanagisawa-Miwa et al. 5 have shown that a single intracoronary infusion of bFGF increased the number of arterioles and capillaries in the infarct border zone after 1 week. The authors observed an apparent reduction in infarct size in animals treated with bFGF and proposed that this reduction resulted from bFGF-induced formation of new microvasculature. In a similar model Unger et al. 2s demonstrated enhanced collateral flow in dogs exposed to daily in-
May 1995
930
Landau, Jacobs, and Haudenschild
tracoronary bFGF for 28 days, also supporting an angiogenic effect of this peptide in myocardium. In a porcine model of microinfarctions induced by intracoronary injection of agarose beads, animals were exposed to bFGF, which was adsorbed to the beads and was presumably released locally. Higher microvessel counts were observed in animals treated with bFGF compared with animals in the control group 2 weeks after infarction. The difference was most marked in sections with extensive evidence of nonviable myocardium, ls As in this study the presence of myocardial necrosis appears to have enhanced the degree of angiogenesis produced by bFGF. Banal et al. 19 applied high doses of sponge-embedded acidic FGF (aFGF) topically to the epicardium in both ischemic and nonischemic dogs and assessed collateral development from an internal mammary artery pedicle placed over the sponge. No new vessel growth was observed, although pronounced smoothmuscle cell proliferation within the wall of arterioles and small arteries in the infarcted territories of animals treated with aFGF was noted. 19 We observed no such phenomenon possibly because of the use of a different growth factor, lower doses, species-related differences, or the fact that tissue fibrosis in this study was patchy. The angiogenic effects of bFGF have also been evaluated in skeletal muscle beds. Chleboun et al. s implanted ethylene-vinyl acetate pellets that were impregnated with bFGF into rat hind limbs rendered ischemic by ligation of the common iliac, common femoral, and superficial femoral arteries. The impregnated pellets released bFGF in vitro for a period of at least 24 hours. After 3 weeks rats treated with bFGF had improved arterial flow measured by a Doppler technique compared with the control group, although the difference became nonsignificant at the 4-week observation point. The authors did not perform any direct analysis of the degree of collateral flow, nor was a histologic analysis undertaken to characterize the mechanism of the improved arterial flow in the bFGF group. Baffour 9 used a similar model in the rabbit, administering i or 3 ttg bFGF via daily intramuscular injections at four sites in the thigh and one site in the calf for a 2-week period. Response to therapy was assessed with transcutaneous oximetry, angiography, and clinical limb status. The authors found a doserelated improvement in these indexes of muscle perfusion and viability in the bFGF group compared with those of the control group. Similar results have also been demonstrated with intramuscular injec-
AmericanHeartJournal
tions of endothelial cell growth factor, an acidic FGF, and heparin. 29 In conclusion, exposure of a variety of tissues to locally delivered bFGF results in new capillary formation in vivo. Moreover the extent of bFGF-induced angiogenesis was sufficient to alter indexes of tissue performance to a physiologically significant degree in some models of vascular insufficiency. 5, 8, 9 These observations suggest that local exposure of muscle beds to bFGF could produce physiologically important effects on tissue perfusion and suggest the capability for salutary effects in ischemic tissues. Conclusions. We have demonstrated the ability of locally administered bFGF to induce new vessel growth when applied within the pericardial space. This series of experiments also suggests that ischemia facilitates bFGF-induced myocardial angiogenesis in this model. Whether the enhanced vascular supply can be sustained after bFGF is no longer provided is an important issue that will require additional study. Study limitations. Although neovascularization was observed in this model, the source of new vessel growth could not be definitively identified with the methods used. The left ventricular hypertrophy model used is presumed to be ischemic; however, the replacement fibrosis observed may be the result of direct angiotensin II myocyte toxicity rather than infarction caused by supply/demand mismatch. This finding does not alter the study findings that in the setting of inadequate capillary supply caused by left ventricular hypertrophy, bFGF efficiently promotes new vessel growth. REFERENCES
1. Canty JM, Klocke FJ. Reduced regional myocardial perfusion in the presence of pharmacologic vasodilator reserve. Circulation 1985;71:370-7. 2. Landau C, Jacobs AK, Haudenschild CC. Left ventricular hypertrophy induced by angiotensin II is accompanied by a dose dependent fibrotic response [Abstract]. Circulation 1992;86[suppl I]:I-754. 3. Yancy CW, Vissing S, Cuckey JC, Bellomo JF, Firth BG, Blomqvist CG. Maximal conductance versus maximal oxygen uptake in patients with congestive heart failure [abstract]. J Am Coll Cardiol 1988;ll[suppl A]:73A. 4. Halperin JL, creager MA. Arterial obstructive diseases of the extremeties. In: Loscalzo J, Creager MA, Dzau VJ, eds. Vascular medicine. Boston: Little, Brown and Company; 1992:835-43. 5. Yanagisawa-Miwa A, Uchida Y, Nakamura F, Tomaru T, Kido H, Kamijo T, Sugimoto T, Kaji K, Utsuyama M, Kurashima C, Ito H. Salvage of infarcted myocardium by angiogenic action of basic fibroblast growth factor. Science 1992;257:1401-3. 6. Chilian WM, Ackell PA. Nonuniform transmural alpha-adrenergic coronary constriction in the presence of a stenosis. Circ Res 1981;48:41623. 7. Martin WH, Montgomery J, Snell PG, Sokolov JJ, Buckey JC, Maloney DA, Blomqvist CG. Cardiovascular adaptations to intense swim training in sedentary middle-aged humans. Circulation 1987;75:323-30.
Volume 129, Number 5 American Heart Journal
8. Chleboun JO, Martins RN, Mitchell CA, Chirfla TV. bFGF enhances the development of collateral circulation after acute arterial occlusion. Biochem Biophys Res Commun 1992;185:510-6. 9. Baffour R, Berman J, Garb JL, Rhee SW, Kaufman J, Friedmann P. Enhanced angiogenesis and growth of collaterals by in vivo administration of recombinant basic fibroblast growth factor in a rabbit model of acute lower limb ischemia: dose-response effect of basic fibroblast growth factor. J Vasc Surg 1992;16:181-91. 10. Gospodarowicz D, Ferrara N, Schweigerer L, Neufeld G. Structural characterization and biological functions of fibroblast growth factor. Endocr Rev 1987;8:95-114. 11. Muller R, Bravo R, Burkshardt J, Curran T. Induction of c los gene and protein by growth factors precedes activation of c myc. Nature 1984;312:716-9. 12. Cuevas P, Gonzalez AM, Carceller F, Baird A. Vascular response to basic fibroblast growth factor when infused into the normal adventitia or into the injured media of the rat carotid artery. Circ Res 1991;69:360-9. 13. Klagsbrun M, Edalman ER. Biological and biochemical properties of fibroblast growth factors: implications for the pathogenesis of atherosclerosis. Arteriosclerosis 1989;9:269-78. 14. Thompson JA, Anderson KD, DiPietro JM, Zwiebel JA, Zametta M, Anderson WF, Maciag T. Site-directed neovessel formation in vivo. Science 1988;241:1349-52. 15. Thompson JA, Handenschild CC, Anderson KD, DiPietro JM, Anderson WF, Maciag T. Heparin binding growth factor 1 induces the formation of organoid neovascular structures in vivo. Proc Natl Acad Sci U S A 1989;86:7928-32. 16. Eppley BL, Doucet M, Connolly DT, Eerier J. Enhancement of anglogenesis by basic fibroblast growth factor in mandibular bone graft healing in the rabbit. J Oral Maxillofac Surg 1988;46:391-8. 17. Hayek A, Culler FL, Beattie GM, Lopez AD, Cuevas P, Baird A. An in vivo model for the study of the angiogenic effects of basic fibrobIast growth factor. Biochem Biophys Res Commun 1987;147:876-80. 18. Battier A, Scheinowitz M, Bor A, Hasdai D, Vered Z, Di Segni E, Varda-BIoom N, Nass D, Engelberg S, Eldar M, Belkin M, Savion N. Intracoronary injection of basic fibroblast growth factor enhances angio-
Landau, Jacobs, and Haudenschild
19.
20.
21.
22. 23.
24.
25. 26. 27.
28.
29.
931
genesis in infarcted swine myocardium. J Am Coll Cardiol 1993;22: 2001-6. Banal S, Jaklitsch MT, Casscells W, Shou M, Shrivasfar S, Correa R, Epstein SE, Unger EF. Effects of acidic fibroblast growth factor on norreal and ischemie myocardium. Circ Res 1991;69:76-85. Rakusan K, Flanagan MD, Geva T, Southern J, Van Praagh R. Morphometry or human coronary capillaries during normal growth and the effect of age in left ventricular pressure-overload hypertrophy. Circulation 1992;86:38-46. Breisch EA, White FC, Nimmo LE, Bloor CM. Cardiac vasculature and flow during pressure-overload hypertrophy. Am J Physiol 1986; 251:H1031-7. Hudlicka O. Growth of capillaries in skeletal and cardiac muscle. Circ Res 1982;50:451-6L Edelman ER, Mathiowitz E, Langer R, Klagsbrun M. Controlled and modulated release of basic fibroblast growth factor. Biomaterials 1991;12:619-26. Brown J J, Chapuis G, Robertson JIS. The effect of prolonged intravenous infusion of angiotensin in the rabbit. Clin Sci (Colch) 1964;26:16575. Tan L, Jalil JE, Rick R, Janicki JS, Weber KT. Cardiac myocyte necrosis induced by angiotensin II. Circulation 1992;69:1185-95. Burgess WH, Maciag T. The heparin-binding (fibroblast) growth factor family of proteins. Annu Rev Biochem 1989;58:575-606. Vlodavsky I, Folkman J, Sullivan R, Fridman R, Ishai-Michaeli R, Sasse J, Klagsbrun M. Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix. Proc Natl Acad Sci U S A 1987;84:2292-6. 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 blood flow in a canine model. Am J Physiol 1994;266:H1588-95. Pu L-Q, Sniderman AD, Brassard R, Lachapelle KJ, Graham AM, Lisbona R, Symes JF. Enhanced revascularization of the ischemic limb by angiogenic therapy. Circulation 1993;88:208-15.