Adenoviral-mediated expression of antisense RNA to basic fibroblast growth factor reduces tangential stress in arterialized vein grafts

AbstractPurpose: The purpose of this study was to test whether basic fibroblast growth factor (bFGF) participates in arterialized vein graft remodeling. Methods: Rabbits underwent in vivo gene transfer and carotid interposition vein grafting. Segments of external jugular vein were infected with an adenovirus that expressed antisense bFGF RNA (Ad.ASbFGF) at 1 × 1010 PFU/mL to inhibit new synthesis of bFGF by cells in the vein graft wall. Control rabbits were treated with either adenovirus that encoded β-galactosidase (Ad.lacZ) at 1 × 1010 PFU/mL or vehicle (phosphate-buffered saline solution [PBS]). At 3 days, 3 grafts per treatment group were harvested for the determination of gene expression of ASbFGF RNA by reverse transcriptase-polymerase chain reaction. Rabbits were killed, and perfusion was fixed 2 months after the grafting. Total wall thickness and lumen circumference of vein grafts and normal arteries were measured in cross sections. Calculated mean tangential stress (±SD) for the ASbFGF-treated group and controls was compared for significance. Grafts were immunohistochemically stained to assess bFGF protein production. Results: Only the grafts infected with the Ad.ASbFGF gene expressed ASbFGF RNA. Grafts that were treated with Ad.ASbFGF displayed lower tangential stress (10.9 ± 2.3 dynes/cm2) than PBS alone (22 ± 2.8 dynes/cm2) or Ad.lacZ-treated controls (20.6 ± 5.4 dynes/cm2; P <.001). Tangential stress in the Ad.ASbFGF group was comparable to a normal carotid artery (13.9 ± 2.1 dynes/cm2). The difference in mean total wall thickness was significant among the 3 treatment groups: Ad.ASbFGF, 164 ± 3.4 μm); Ad.lacZ, 100 ± 3.3 μm; and PBS, 96 ± 3.6 μm; P <.01). Luminal circumference was not different among the groups. The Ad.ASbFGF-treated vein graft wall was composed of thick layers of concentric smooth muscle cells and elastin fibers in contrast to the sponge-like appearance observed in control arterialized vein grafts. Reduction in bFGF protein was noted only in the Ad.ASbFGF-treated group. Conclusion: Inhibition of bFGF synthesis in vivo with the use of adenoviral gene transfer of antisense RNA to bFGF promotes a vein graft with decreased tangential stress while maintaining the luminal area. The vein graft wall is remodeled and qualitatively resembles an artery so that wall tangential stress in Ad.ASbFGF and normal artery are not significantly different. The lack of significant difference in lumen circumference among groups suggests that wall thickening in the Ad.ASbFGF grafts is not at the expense of luminal narrowing. Our results suggest that ASbFGF RNA expression may represent an effective strategy in limiting the failure of arterialized venous conduits. (J Vasc Surg 2000;31:770-80.