epithelium, it would be much safer to have this occur in the nose than 
in the bronchus. Finally, the nasal epithelium can be sampled with 
greater ease, safety and frequency than the bronchial epithelium. 
(2) Although it is safer and easier to evaluate the vector in the nasal 
epithelium, the dominant clinical manifestations are between the 
larynx and the terminal bronchioles i.e., correction of the biologic 
abnormality of CF in the nasal epithelium does not offer the patient 
any possibility of correction of the fatal manifestation of the 
disease. Based on the underlying concept that there are theoretical 
inherent risks in administering a replication deficient recombinant 
adenovirus to humans, the protocol must offer possible efficacy to the 
patient. In this regard, it should include administration of the 
vector to the bronchial epithelium. 
(3) There are a limited number of sites in the respiratory epithelium that 
can be accessed (right nostril, left nostril, right large bronchi, 
left large bronchi) for assessment of biologic efficacy (e.g., 
epithelial cells removed, potential difference measured) . 
(4) The epithelium can be obtained by brushing with a fine cytology brush. 
However, there is a limitation in the frequency of times the same site 
can be evaluated; at a minimum it has to be assumed that it will take 
7 days to "heal" the nasal epithelium after sampling. Another 
consideration is the difference in accessibility of the nose and 
bronchi; whereas the nasal epithelium can be readily sampled, there is 
a limitation in the frequency of bronchoscopies that can be done to 
sample the bronchial epithelium in individuals with CF. 
(5) Maximal information will be derived in the protocol by having as many 
controls as possible. Since the patients will serve as their own 
controls, this can be achieved in two ways: (a) by having a baseline 
period to assess various parameters before administration of the 
vector; (b) by administration of the "vehicle" used in the vector 
preparation to assess the effect of the vehicle on various parameters; 
and (c) by administration of the vector to one side (e.g. right versus 
left nostril, right versus left large bronchi). 
(6) There are anatomic difficulties in reaching the small bronchi compared 
to the large airways . Further, if there are adverse effects in the 
bronchi, it is preferable to have these in the large airways for both 
accessibility and management of adverse physiologic consequences. 
Conclusions Regarding Design of the Protocol 
(1) Administer the vector first to the nasal epithelium, then to the large 
bronchi epithelium of the same individual. Administration to the nasal 
epithelium first will permit assessment of acute toxicity in a site 
with a minimal potential for airway compromise and will permit 
frequent assessment of biologic efficacy; administration to the large 
bronchi epithelium has the advantages of the possibility of clinical 
efficacy as well as assessment of biologic efficacy in airways, and is 
[732] 
Recombinant DNA Research, Volume 16 
