M.J. Welsh and A.E. Smith, RAC Application 
Potential risks and adverse effects of the protocol include the following. 
a. Risks associated with inflammation, cytopathic effects, and immune response. 
It is possible that inflammation locally at the site of application of the virus could occur. 
Local cytopathic effects on the respiratory nasal epithelium are also possible. Inflammation 
or cytopathic effects include local pain, bleeding or subsequent scarring; although serious 
effects appear very unlikely. The risk is minimized by the use of small doses of the virus in 
a very limited area of nasal mucosa. The use of nasal mucosa, rather than bronchial 
mucosa, is a major safety feature that should allow us to assess safety without endangering 
the patient. Our animal studies, which used much higher doses of virus, indicate that if 
inflammation should occur, it will be mild and transient. Our studies with human airway 
epithelia suggest that there will be no cytopathic effects. 
It is possible that the administration of Ad2/CFTR-1 could produce an immune response to 
the recombinant virus that was not present before the study. While it is unlikely that this 
would affect the patient during the study, the presence of immune sensitization could affect 
the response to subsequent treatment with adenovirus vectors. However, the risk to the 
participants of this study will be minimized because they will already be seropositive. 
These considerations may be more of an issue in seronegative patients or with repeated 
administration of the virus. 
b. Risks associated with virus replication. 
The Ad2/CFTR-1 viral construct has been rendered defective for replication by deletion of 
two important early genes El a and Elb. However, it is possible that the virus will have a 
limited ability to replicate in human cells. Furthermore, under certain circumstances, the 
viral defect could be complemented. Such circumstances include low level contamination 
of the Ad2/CFTR- 1 preparation with wild-type virus, coinfection with wild-type 
adenovirus, or provision of El gene function by latent or residual adenovirus resulting from 
an earlier infection (100), or by infection by other viruses able to provide such function 
(127-129). Finally, epithelia might also provide normal cellular proteins with El-like 
functions that are able to complement the defective virus (101,102). 
Most of these possibilities seem remote. The Ad2/CFTR-1 virus is defective not only 
because it lacks El, but also because it is so large as to be difficult to package. Thus, even 
in permissive cells that provide El functions, virus replication is modest relative to wild- 
type. Second, the likelihood of coinfection of treated cells with wild-type virus is 
minimized by prior screening of the patients, by treating only seropositive patients, by 
treating only a small area of nasal epithelia and by isolation of the patient during treatment. 
Third, coinfection with another virus, although possibly providing some El activities, is 
unlikely to complement all El gene functions (145). Fourth, in the unlikely event that 
coinfection or contamination with wild-type adenovirus did occur, the wild-type virus 
would probably compete out the defective virus resulting, eventually, in a wild-type 
infection that would be expected to become self-limited. Model experiments, B2a(4), 
confirm that wild-type virus rapidly overgrows Ad2/CFTR-1. Because there is data to 
suggest the presence of El sequences from previous adenovirus infections in some 
individuals (100), we will screen patient's nasal cells for El DNA by PCR and exclude 
those patients with detectable signal. 
There is also a possibility of recombination between wild-type virus and the Ad2/CFTR-1 
virus. Whether this occurred by legitimate or illegitimate recombination, for any progeny 
virus to be viable would require two events to occur. First, insertion of the missing El 
Recombinant DNA Research, Volume 16 
[915] 
