Safety of retrovirus gene therapy, Temin, 
competent retrovirus. It is known that VL30 and MLV sequences have recombined and 
formed recombinant proviruses (Itin and Keshet, 1983). The likelihood of such recom- 
bination in human cells is unknown, but it probably is better to insist that vector prepara- 
tions to be used for human infection are not contaminated with VL30-like sequences by 
growing the vector in a cell not expressing such sequences. 
Finally, we consider the probability of insertion of the vector in the patient’s 
germ-line. Since the vector is so severely crippled and the infections with the vector will 
be performed outside the body with helper virus-free stocks, this possibility seems to be 
remote. Furthermore, infection of the germ-line is unlikely to lead to any effects. Most 
infecting retroviruses are not transcribed in the germ-line (see references in Temin, 
1986b). The integration could be mutagenic, but there are already over 500,000 copies of 
retrosequences in the germ-line without observable effects upon viability (Temin,1985). 
In addition, the human retroviruses, HIV, HTLV-I, and HTLV-II have not become 
endogenous (Samgadharan et al.,1985). Thus, even if germ-line integration occurred, it 
would not have observable consequences. 
El. SOCIAL PROBLEMS (INDIVIDUAL AND GENERAL) 
A. When Should Retrovirus Gene Therapy Be Used? 
Somatic gene therapy of human disease with retrovirus vectors is a technique 
designed to correct single gene human diseases. A retrovirus vector delivers to cells that 
are defective for genes-cneodia g e nzymes or o th e r prot ei n s- gen es that can form the active 
functioning enzymes or other proteins to be used in that cell or exported from that cell. 
As such, the determination of which patients should receive this therapy is the same as for 
any experimental treatment. However, since in some respects retrovirus gene therapy is 
so novel in methodology compared to other new therapies and since there is some pos- 
sibility of inducing later cancers, the selection criteria should be very stringent. Thus, ini- 
Recombinant DNA Research, Volume 12 
[ 386 ] 
