humans. Type 5 adenovirus is categorized in subgroup C, a subgroup that is 
nontumorgenic in experimental animals. Further, there is extensive clinical 
experience regarding the administration of live adenoviruses to humans with 
no reports of association with malignancy. 
With this background, it is reasonable to conclude that there will be no 
risk for malignancy associated with this vector that will be discernable 
above the background risk. 
4.11 Does the Genome of the Vector Integrate into the Genome of the Target 
Cells? 
The available data suggests that adenoviruses do not integrate their genome 
into the genome of the target cells (Karlsson et al., 1985; Karlsson et 
al . , 1986). It is not possible to evaluate this with freshly isolated 
airway epithelial cells from normals or individuals with CF (too few cells, 
difficulty in maintaining in culture, inability to have the cells replicate 
to capitalize on cloning single cells with selectable marker) . The frequen- 
cy of adenovirus integration is estimated to be very low (Karlsson et al., 
1985; Karlsson et al . , 1986). 
In the context of the above, it is likely that the risk for insertional 
mutagenesis from AdCFTR is extremely low, far lower than that with 
retroviruses . 
4.12 Does the Expression of CFTR Need to be Regulated? 
AdCFTR and AvlCFl are designed with a constitutive adenovirus promot- 
er/enhancer to drive the expression of the human CFTR cDNA. In the human 
airway, CFTR gene expression is very low (Trapnell et al . , 1991a), with 
somewhat higher expression in glandular serous cells than in the surface 
epithelial cells ( J . Wilson, personal communication). The promoter has a 
number of putative transcriptional control regions and in vitro studies 
have demonstrated that inflammatory stimuli will down-regulate expression 
of the genes (Yoshimura et al . , 1991; Chou et al . , 1991; Trapnell et al . , 
1991b; Bargon et al . , 1992). 
From the safety viewpoint, since the mutations of the CFTR gene cause the 
airway epithelial cells to be "deficient" in CFTR function, the major 
safety issue regarding control of expression is to ask what are the conse- 
quences of over-expression. Three lines of evidence suggest it is not a 
concern. 
First, in the surface epithelium of the normal lung, endogenous CFTR gene 
expression cannot be detected with the methods currently used. In cotton 
rats treated with Ad-CFTR by intratracheal administration, human CFTR gene 
expression is readily detectable in the surface epithelium, i.e., there is 
likely over-expression of the CFTR gene (see Appendix 1). This has no 
clinical consequences to the animals (see Appendix 1 and Section 4.4). 
Second, transgenic mice cons titutively expressing the human CFTR cDNA in 
all organs (i.e., with a promoter used by all cells) are healthy (Whitsett 
Recombinant DNA Research, Volume 16 
