Safety of direct myocardial administration of an adenovirus vector encoding vascular endothelial growth factor 121

A gene therapy strategy involving direct myocardial administration of an adenovirus (Ad) vector encoding the vascular endothelial growth factor 121 cDNA (Ad(GV)VEGF121.10) has been shown to be capable of "biological revascularization" of ischemic myocardium in an established porcine model [Mack, C.A. (1998). J. Thorac. Cardiovasc. Surg. 115, 168-177]. The present study evaluates the local and systemic safety of this therapy in this porcine ischemia model and in normal mice. Myocardial ischemia was induced in Yorkshire swine with an ameroid constrictor 21 days prior to vector administration. Ad(GV)VEGF121.10 (10(9) or 10(10) PFU), Ad5 wild type (10(9) PFU), AdNull (control vector with no transgene; 10(9) PFU), saline, or no injection (naive) was administered in 10 sites in the ischemic, circumflex distribution of the myocardium. Toxicity was assessed by survival, serial echocardiography, blood analyses, and myocardial and liver histology at 3 and 28 days after vector administration. All pigs survived to sacrifice, except for one animal in the Ad(GV)VEGF121.10 (10(10) PFU) group, which died as a result of oversedation. Echocardiograms of Ad(GV)VEGF121.10-treated pigs demonstrated no differences in pericardial effusion, mitral valve regurgitation, or regional wall motion compared with control pigs. Intramyocardial administration of Ad(GV)VEGF121.10 included only minimal myocardial inflammation and necrosis, and no hepatic inflammation or necrosis. Only a mild elevation of the white blood cell count was encountered on day 3, which was transient and self-limited in the Ad(GV)VEGF121.10 group as compared with the saline-treated animals. As a measure of inadvertent intravascular administration of vector, normal C57/BL6 mice received intravenous Ad(GV)VEGF121.10 (10(4), 10(6), 5 x 10(7), or 10(9) PFU), AdNull (5 x 10(7) or 10(9) PFU), or saline. Toxicity was assessed by survival, blood analyses, and organ histology at 3 and 7 days after vector administration. A separate group of C57/BL6 mice received intravenous AdmVEGF164 (Ad vector encoding the murine VEGF164 cDNA), Ad(GV)VEGF121.10, AdNull (10(8) PFU each group), or saline to assess duration of expression and safety of a homologous transgene. All mice survived to sacrifice except for 40% of the mice in the highest (10(9) PFU; a dose more than 10(3)-fold higher by body weight than the efficacious dose in pigs) Ad(GV)VEGF121.10 dose group, which died on days 5-6 after vector administration. The only differences seen in the blood analyses between treated and control mice were in the very high Ad(GV)VEGF121.10 dose group (10(9) PFU), which demonstrated an anemia as well as an increase in alkaline phosphatase when compared with all other treatment groups. Hepatic VEGF levels by ELISA in AdmVEGF164-treated mice did not persist beyond 14 days after vector administration, suggesting that persistent expression of a homologous VEGF gene transferred with an Ad vector is not a significant safety risk. Although this is not a chronic toxicity study, these data demonstrate the safety of direct myocardial administration of Ad(GV)VEGF121.10, and support the potential use of this strategy to treat human myocardial ischemia.