C. Risks of Gene Transfer and Gene Expression 
The only possible complication of LDL receptor gene transfer per se would be an immune 
response to the recombinant derived LDL receptor protein. Issues that relate to this potential 
problem have been described under Section III.H. It is difficult to predict the likelihood of this 
occurring, however, our work in the "receptor negative" WHHL rabbit (Appendix G) suggests 
that immunologic rejection may not be a confounding or limiting problem. The most morbid 
consequence of an immune response to the recombinant derived receptor would simply be loss of 
functional engraftment. 
Risks of gene transfer using retroviral vectors have been addressed extensively in 
animal models and more recently in human subjects (Reviewed in Ref. 121). Possible 
complications include 1) the development of replication competent virus and sequalae of viral 
replication in vivo, and 2) long-term consequences of random retroviral integration such as the 
development of secondary malignancies. 
A significant effort has been made to minimize the risks of formation of replication 
competent viruses. The cell line that produces the virus has been engineered to prevent the 
formation of wild-type virus (See Section C.). The potential toxicity of inadvertent formation 
of replication competent murine retroviruses has been addressed in a series of experiments 
performed at the NIH (121, 122). Four nonhuman primates were infused with large volumes 
of replication competent amphotropic virus. The virus was rapidly cleared and the animals 
remained free of disease up to the time of their report (mean follow-up was 43.9 months). 
Additional experiments in which monkeys were immunosuppressed and transplanted with virus 
producing autologous fibroblasts were associated with no ontoward effects for the average 
follow-up of 61 .8 months. Based on these extensive studies we conclude that the formation of 
replication competent virus is highly unlikely and if it occurred would probably not cause 
disease. 
A more difficult safety issue to address is the risk of malignant transformation due to 
retroviral mediated gene transfer. The theoretical basis for this has been nicely reviewed by 
Cornetta et al. (121) The major concern relates to the fact that retroviruses insert into the 
recipient cell's genome in a near-random manner and that this insertion event could potentially 
lead to the activation of an oncogene or the inactivation of a tumor suppressor gene. While this 
is theoretically possible the chance of a secondary malignancy is highly unlikely for several 
reasons. Transformation of human cells in vivo is clearly a multistep process and a single 
event, such as insertion of a provirus, will probably not lead to the formation of a de novo 
tumor. Furthermore, insertional inactivation of a tumor suppressor gene would be insufficient 
for transformation because one allele would remain intact and capable of providing tumor 
suppression function. While the insertion may not itself cause transformation it may bring the 
cell one step closer to transformation. Applications of retrovirus mediated gene transfer for 
treatment of humans must take into account this potential tumor promoting capability. We 
believe this will not be a major problem in our protocol for several reasons. We will be 
working with a population of patients that are relatively young and have a short life expectancy. 
Development of secondary malignancies long after the therapy will not be an issue. 
Furthermore the patient's tumor surveillance systems will remain intact because they will not 
be immunosuppressed. 
Another aspect of our protocol that makes transformation highly unlikely relates to the 
nature of the recipient cell for gene transfer, the hepatocyte. Under normal conditions, the 
hepatocyte is a terminally differentiated cell which is capable of undergoing very limited 
number of cell divisions in vivo or in vitro. It is extremely difficult to immortalize or 
[194] 
Recombinant DNA Research, Volume 15 
