did in fact correspond to the region between the two primers. Direct sequencing of the RT-PCR product 
documented that this band represented the human VEGF sequence. 
Development of collateral vessels in the ischemic limb was serially evaluated by calf blood 
pressure measurement and internal iliac arteriography immediately prior to transfection (day 0), and then in 
serial fashion at days 10 and 30 post-transfection. Following the final 30-day follow-up, the animal was 
sacrificed, and tissue sections were prepared from the hindlimb muscles in order to perform analysis of 
capillary density. 
The development of collateral vessels in the 5 rabbits transfected with phVEGFi 65 and 6 rabbits 
transfected with pGSVLacZ was evaluated by selective internal iliac angiography. In control animals, 
collateral artery development in the medial thigh typically appeared unchanged or progressed only slightly 
in serial angiograms recorded at days 0, 10, and 30. In contrast, in the VEGF-transfected group, marked 
progression of collateral artery was observed between days 10 and 30. Morphometric analysis of 
collateral vessel development in the medial thigh was performed by calculating the angiographic score. At 
baseline (day 0), there was no significant difference in angiographic score between the VEGF-transfected 
and control groups (day 0: 0.17±0.02 vs 0.20±0.06, p=ns). By day 30, however, the angiographic score 
in VEGF-transfected group was significantly higher than that in control group (0.47±0.09 vs 0.34±0.10, 
p<0.05). 
Reduction of the hemodynamic deficit in the ischemic limb following VEGF-transfection was 
confirmed by measurement of calf blood pressure ratio (ischemic/normal limb). The calf blood presssure 
ratio was virtually identical in both groups prior to transfection (0.23±0.14 in control and 0.20±0.12 in 
VEGF-transfected animals, p=ns). By day 10 post-transfection, the blood pressure ratio for the VEGF- 
transfected rabbits was significantly higher than for the control rabbits (0.60±0.12 vs 0.32±0.14, 
p<0.01). At day 30, the blood pressure ratio for VEGF-transfected group continued to exceed that of 
controls (0.70±0.08 vs 0.50±0.18, p<0.05). 
A favorable effect of VEGF-transfection upon revascularization was also apparent at the 
capillary level. The medial thigh muscles of the ischemic limbs were histologically examined at day 30 
post-transfection. Analysis of capillary density disclosed a value of 233.0±60.9 /mm 2 in VEGF- 
transfected group versus 168.7±31.5 /mm 2 in the control group (p<0.05). Analysis of capillary/myocyte 
ratio disclosed a value of 0.67±0.15 in the VEGF-transfected group versus 0.48+0.10 in the control group 
(p<0.05). 
To evaluate the efficiency of in vivo arterial gene transfer, transfected iliac arteries were 
harvested at 5 days post-transfection, and were used for B-galactosidase histochemical analysis. In arteries 
transfected with nuclear targeted B-galactosidase, evidence of successful transfection, indicated by dark 
blue nuclear staining, was observed in only < 0.5 % of total arterial cells. Arteries transfected with 
phVEGF ]65 were negative for nuclear staining. 
This in vitro and in vivo data is presented in further detail in a manuscript submitted for 
publication ("Therapeutic Angiogenesis Following Arterial Gene Transfer of Vascular Endothelial Growth 
Factor in a Rabbit Model of Hindlimb Ischemia") which is enclosed in the Appendix to this proposal. 
More recently, we have evaluated the physiologic response to VEGF gene transfer in the same 
rabbit model of hindlimb ischemia 39. At 30 days post-transfection, limb perfusion was evaluated pre- and 
post-papaverine (2 mg, intra-arterial) using a .018 in. guidewire with a Doppler crystal at the distal tip 
38to measure flow velocity, and quantitative angiography to measure iliac artery diameter at the site of wire 
placement. Maximum flow (ml/min post-papaverine), calculated assuming circular lumen geometry, was 
62.3+6.6 (n=7) following gene therapy, similar to that recorded following 500 |ig of recombinant protein 
in the same animal model (52.7 ±2.4, n=ll), and significantly greater than that recorded in untreated 
controls (30.9+2.6, n=6, p<0.001 by ANOVA). 
The experimental model used for these pre-clinical studies was designed to simulate ischemia 
characteristic of patients with lower extremity arterial occlusive disease. In contrast to models of acute 
ischemic injury, an interval of 10 days between the time of surgery and gene transfer was incorporated to 
allow for development of varying degrees of spontaneous collateral formation. The notion that this animal 
model simulates "chronic" ischemia is further supported by the finding that collateral vessel development 
(in untreated controls) does not improve significantly beyond this point, up to the time of final 
angiography and death at 40 days 7. Indeed, initial studies evaluating this model 8 demonstrated no 
significant "spontaneous" improvement in collateral perfusion either angiographically or 
hemodynamically between days 30 and 90 after surgically induced ischemia. 
Our pre-clinical studies thus suggest that VEGF arterial gene transfer can successfully 
augment collateral perfusion despite a low transfection efficiency. On the basis of control animals 
transfected with pGSVLacZ, we estimate that the transfection efficiency achieved with the current delivery 
Recombinant DNA Research, Volume 20 
[323] 
