Recombinant DNA Advisory Committee - 10/7-8/91 
respect to the issue of safety, he asked why the investigators had not inserted the TK 
gene into the tumor cells that could then be killed by ganciclovir if necessary. 
Dr. Carmen recounted the research activities that Dr. Rosenberg has presented to the 
RAC over the past several years. Dr. Carmen described the two protocols currently 
under the committee's consideration. With respect to these proposals, he noted that 
there is evidence that subcutaneous injections will be more effective than parenteral 
injections, There is evidence that the co-injection of modified and unmodified tumor 
cells can lead to the inhibition of growth for both types of cells. There is also evidence 
that IL-2 has much in common with TNF, both are known cytokines and can precipitate 
serious side effects that are treatable. While the reintroduction of modified tumor cells 
could cause tumor growth, appropriate counter measures are available if the patients 
have failed all other possible avenues of treatment. 
Dr. Carmen asked what criteria Dr. Rosenberg will use to determine which patients will 
receive the gene for TNF and which patients will receive the gene for IL-2. He noted 
the NIH IBC had recommended a pilot study which would test the subcutaneous 
insertion of autologous tumor, putting aside the recombinant DNA phase of the protocol. 
The HGTS considered such studies unnecessary by vetoing the insertion of unmodified 
tumor cell lines and substituting its own safety measures. Using only cells that were 
transduced with cytokine genes in these experiments seems to be a drastic revision. 
Mr. Barton asked if it is possible to administer TNF together with the tumor cells, rather 
than using the TNF-modified tumor cells to control the dosage of the TNF more 
precisely. He questioned whether the TNF and the IU2 protocols are actually as 
parallel as one might expect, given the significantly greater toxicity of the TNF. The 
number of patients seems high considering the uncertainties. In the consent form, it 
should be made clear how the patients are to be involved in the cost sharing. With 
regard to risks, it should be noted that once the genetically engineered tumor cells have 
been administered to the patients, the dosages of IL-2 or TNF may not be controllable. 
The hypothesis should be clearly explained in the consent form as well as how the 
treatment might be a positive step forward. 
Dr. Mclvor responded to Dr. Carmen's concerns that injection of transduced cells alone, 
versus transduced cells combined with untransduced cells. In animal models, only 
transduced cells elicited an increased immune response. Therefore, we must assume that 
they are expressing IL-2 or TNF at increased levels which stimulates this immune 
response and an anti-tumor response against these cells. The injection of untransduced 
cells along with transduced cells would be much riskier and may not stimulate the same 
response. This is the logic behind eliminating the co-injection of untransduced cells. 
Mr. Capron asked the investigators to explain an aspect of the consent document 
[ 22 ] 
Recombinant DNA Research, Volume 15 
