3 
all of the gag , pol , and env genes are deleted and the gene(s) of interest 
inserted. The major differences between the various research groups currently 
developing vectors for human gene therapy are the modifications introduced 
into these regions. 
Dr. Anderson said investigators attempting to engineer retroviral vectors face 
several problems. The major problem is constructing a "workable" vector. The 
vector must be introduced into cells of the animals' body, and the gene of 
interest expressed at reasonable levels. At this time, high expression of 
these vectors has not been obtained in animals. Vectors which should provide 
improved expression are being developed, but it is not yet known vhether these 
vectors will have high enough expression levels to be useful in human gene 
therapy. High levels of expression depend on the control regions and to a 
lesser extent the introduced gene. 
Dr. Anderson said the second issue faced by investigators is whether unantici- 
pated untoward effects may occur. The important considerations in evaluating 
the possibility of untoward effects are described in detail in the "Points to 
Consider in the Design and Submission of Human Somatic Cell Gene Therapy Protocols" 
developed by the working group. 
Mr. Capron asked whether untoward events were possible with retroviral vectors 
since basic retroviral functions are not well understood. He asked Dr. Anderson 
to offer a hypothetical scenario of such an event. Dr. Anderson said a primary 
conjectural effect would be induction of cancer through insertional activation 
of cellular proto-oncogenes. 
Dr. Anderson reminded the group that gene therapy would be applied to seriously 
ill individuals, and risk/benefit considerations would have to be balanced. 
He said risk is inherent in any procedure. He offered the example of Cavid 
the "bubble boy." David, who was born with severe combined immune deficiency 
disease, received a bone marrow tranplant from his sister in the hope that the 
transplant would reconstitute his immune system. However, David died as a 
result of the unexpected activation of latent EBV virus in the bone marrow cells 
of the donor. 
Dr. Anderson said data generated with animal systems could be used to predict a 
probability of inducing carcinogenesis through retrovirus mediated insertional 
activation. He said the risk of injury is expected to be low based on current 
experience with animal systems. 
Dr. George Scangos of Johns Hopkins University said the important issua of 
insertional activation arises because the site of retrovirus insertion cannot 
be controlled, and integration could result in activation of a cellular proto- 
oncogene. Wiether the patient's inmune system would be able to control a 
tumor resulting from activation of a proto-oncogene is not known. 
[ 569 ] 
