Appropriate control of ex vivo gene therapy delivering basic fibroblast growth factor promotes successful and safe development of collateral vessels in rabbit model of hind limb ischemia.

Purpose: In our previous study, adenovirus-mediated ex vivo gene transfer of basic fibroblast growth factor promoted significant collateral vessel development in a rabbit model of hind limb ischemia. The present study examined how to control the efficacy and safety of this gene therapy, and also evaluated the feasibility of repeat application of this procedure.

Methods: Modified hFGF gene with the secretory signal sequence was adenovirally transferred to cultured autologous fibroblasts, and various numbers of the cells (2 x 10(5), 1 x 10(6), 5 x 10(6), or 2.5 x 10(7)) or vehicle was injected through the left internal iliac artery in rabbits in whom the left femoral artery had been excised 21 days previously. Twenty-eight days after cell administration, calf blood pressure ratio, angiographic score, blood flow in the internal iliac artery, and capillary density of muscle tissue were measured to analyze collateral vessel development and tissue perfusion in the ischemic limb. To assess delivery efficiency and viral contamination, the distribution of injected cells and the time course of blood anti-adenovirus antibody titer were examined in rabbits treated with various numbers of gene-transduced cells. In addition, animals received two injections, 21 days apart, of fibroblasts infected with adenovirus vector containing the luciferase gene, and luciferase expression was measured to evaluate whether the present therapy is repeatable.

Results: At 28 days after cell administration, significant collateral vessel development without detectable side effects was observed in rabbits who received 5 x 10(6) or 2.5 x 10(7) cells, compared with those who received vehicle, and no significant development was detected in animals with fewer than 5 x 10(6) cells (P <.01 for calf blood pressure ratio and capillary density, P <.05 for angiographic score and maximum blood flow). There was no difference in collateral augmentation between rabbits with 5 x 10(6) and 2.5 x 10(7) cells. However, in animals with 2.5 x 10(7) cells a large number of injected cells accumulated in the lungs, anti-adenovirus antibody titer increased significantly, and calf blood pressure in the left hind limb of two rabbits decreased immediately after injection. Luciferase analysis showed very low gene expression after repeated administration.

Conclusion: These findings suggest that 5 x 10(6) is a suitable number of cells to induce appropriate collateral vessel development and minimize potential side effects of this procedure. Despite use of ex vivo gene transfer, repeat administration of the cells was not feasible. Clinical relevance Since the present study determined the appropriate conditions for effective and safe stimulation of collateral vessels, the clinical relevance of the ex vivo therapy might be carried forward. However, the findings raised another issue that should be resolved before clinical application; that is, the number of gene-transduced cells able to be injected was strictly limited. To estimate the therapeutic range of cell number in humans, additional experiments using large animals are desirable.