Protocol HNS 94-001 
October 4, 1994 
Page 7 
Table 1 Effect of Ad5CMV-p53 on Established tumor growth In nude mice 8 
Treatment 
Mean volume 
[mm 3 ± SEM] 
Tu-138 (7) Tu-177 (6) 
Ad5CMV-p53 
22.3 ± 14 
13 ± 18 
Ad5(dl312) 
803 ± 300 
533 ± 148 
Medium 
1297 ± 511 
421 ± 143 
Significance 
p value 
p value 
p53 b : dl312 
0.03 
0.02 
p53 : Medium 
0.04 
0.03 
a The cells were injected subcutaneously at 5 x 1 0 6 cells/flap. Tumor sizes were determined at dav 20 
after treatment. Numbers in parentheses represent the number of animals evaluated. 
b Ad5CMVp53 is abbreviated as p53; d!312 is an abbreviation for Ad5(dl312). Statistical anallysis by 
Friedman's two-way anallysis of variance. 
The efficacy of Ad5CMV-p53 in inhibiting tumorigenicity was further evaluated in the mouse 
model of microscopic residual disease in HNSCC. Representative HNSCC cell-lines of 
homozygous mutations, as well as wild-type p53 cell lines were used. In head and neck 
cancer, as well as several other solid malignancies, direct gene transfer to microscopic residual 
carcinoma may not be so technically difficult. When the primary source of tumor is removed, 
the tumor base is readily available for molecular therapy as well as the most likely pathway of 
lymphatic spread when a neck lymph node dissection is performed. To investigate these 
issues, we designe&our experiments to determine if in vivo adenovirus p53 mediated gene 
transfer would effect the establishment or growth of SCCHN cells implanted into a 
subcutaneous flap. Athymic nude female mice were anesthetized and three subcutaneous 
flaps elevated and the SCCHN cell lines pipetted subcutaneously in order to prevent 
erroneous tumor inoculation and dispense a specified number of cells. Following 48 hours, 
mice were reinoculated with either adenovirus p53, transport medium alone, and a non- 
marker replication defective adenovirus dl312 or the adenovirus 6-galactosidase vector. The 
development of tumors was tumor cell number dependent, allotted time for implantation 
dependent, and dose dependent upon adenovirus p53. A representative viral dose 
response experiment is shown in figure 8, Appendix D. Dose response experiments with the 
marker Galactosidase gene clearly demonstrate dose-response transduction efficiency in this 
model (figure 8) and was also confirmed by p53 immunohistochemistry three days following 
infection (figure 9, Appendix D). In the presence of microscopically implanted tumor of 2.5- 
5.0 x 1 0 6 cells treated with adenovirus p53 at 1 0 8 P.F.U. or greater, tumors developed in 2 of 
6 animals in only one of the wild-type cell lines (886LN). All other tumor cell line exhibited total 
inhibition of tumor development both grossly and histologically. These experiments clearly 
indicate that microscopic proliferating tumor cells can be successfully infected in vivo if 
exposed to adenovirus p53. Tumor formation was evaluated at the end of a 8-week period by 
gross and histologic analysis of the surgical sites. The data of tumor measurements are 
summarized in Table II. These results indicate that Ad5CMV-p53 can prevent the formation of 
HNSCC in a subcutaneous model of microscopic residual disease. 
[652] 
Recombinant DNA Research, Volume 20 
