Host responses and persistence of vector genome following intrabronchial administration of an E1(-)E3(-) adenovirus gene transfer vector to normal individuals.

Adenovirus (Ad)-mediated gene transfer to the respiratory epithelium of experimental animals and to nasal and airway epithelium of individuals with cystic fibrosis is followed by transient gene expression. Extensive studies in experimental animals are consistent with the concept that local cellular host anti-vector immune responses account for this short-term expression, and systemic and local [lung epithelial lining fluid (ELF)] anti-Ad neutralizing antibodies are generated following Ad vector administration to the respiratory epithelial surface. To determine if this paradigm holds in normal humans, a first-generation Ad vector (Ad(GV)CD.10, an E1(-)E3(-) Ad serotype 5-based vector coding for the Escherichia coli cytosine deaminase gene) was sprayed locally in escalating doses (8 x 10(8)-8 x 10(10) particle units (pu), n = 2/group) into the lung airway epithelium of six normal individuals. Serum, ELF, and endobronchial biopsies were obtained at baseline and at various time points following vector administration. In contrast to the observations in experimental animals in which lung administration of first-generation Ad vectors is followed by strong systemic and local host response, bronchial spray administration of the Ad vector to normal humans showed: (1) minimal inflammation in bronchial biopsies, bronchial brushing, and bronchoalveolar lavage fluid; (2) no blood lymphocyte proliferation in five of six individuals in response to in vitro stimulation with Ad antigens; and (3) no significant increase from baseline in blood or lung ELF anti-Ad neutralizing antibodies. Despite this minimal normal human anti-Ad host response, dose-dependent levels of vector DNA in the airway epithelium were transient. Vector DNA in the targeted airway epithelial cells peaked in a dose-dependent fashion at 0.007 to 1.1 copies/cell at day 7 and declined thereafter, reducing to <10% of peak levels by 2 weeks. These observations demonstrate both the strengths and the limits of using experimental animals to predict human responses to gene transfer vectors. While the transient nature of Ad vector persistence in the airway epithelium is predicted by most experimental animal studies, respiratory epithelial administration of first-generation Ad vectors at doses up to 10(10) pu to airway epithelium of healthy individuals elicits minimal to no detectable systemic and mucosal humoral and cellular immune responses, an observation diametrically opposed to the host responses measured in experimental animals. These findings suggest that, while adaptive anti-Ad immune responses likely play some role in the disappearance of the vector DNA following vector administration to the human lung, other mechanisms may also be involved in the response of humans to Ad gene transfer vectors.