Human Gene Therapy Subcommittee - July 29-30, 1991 
Dr. Leventhal noted that the protocol stated that the response correlated with the 
amount of TNF produced and could be blocked with anti-TNF antibody. She asked 
whether the amount of TNF produced in vitro correlated with what was observed 
locally. Since tumor growth at the injection site is necessary for TIL production, it 
seems problematic that the TNF-modified tumor cells will not grow. Does the animal 
model suggest how long the cells have to be present to generate the TIL? 
Dr. Leventhal said the consent form should cover the possibility that, if a patient's cell 
line cannot be started or transduced, the patient will not be treated. There was also 
concern about whether only patients with measurable disease would be eligible for 
these studies since the studies are similar to a Phase I/II study. The studies for 
measurement of disease, other than disease that was palpable on physical examination, 
were not completely described. There needs to be more detail as to how the toxicity is 
to be reported There is a question if the patients are going to live long enough for the 
analysis. 
Dr. Rosenberg responded to the reviewers. He noted that the protocols that he has 
begun in collaboration with Dr. Anderson and Dr. Blaese, had involved gene insertion 
into TIL. In the first protocol, the neomycin resistance gene was inserted. Since that 
time, the investigators have made attempts to insert genes to improve the therapeutic 
efficacy of the TIL. The results were described in studies where the TNF gene was 
inserted into the TIL. There were no safety problems and no toxicity due to the TNF 
insertion. The patients receive the cells as out-patients. There has been extensive 
safety testing of the producer cell lines and the infused TIL. 
Dr. Rosenberg said that in the currently proposed protocols, the tumor cells and not 
the TIL will be modified. Relevant experiments from the literature and preclinical 
animal experiments performed in his laboratory were described. The TNF gene was 
introduced into a variety of mouse tumor cells to show that there are clones With 
different levels of TNF production. Unmodified cells do not produce detectable 
amounts of TNF and IL-2. When the unmodified tumor cells were injected into 
animals, the tumor grew. When the TNF-modified tumor cells were injected, the 
tumor grew for ten days and then spontaneously regressed. An antibody to TNF 
abrogates the effect, showing that TNF secretion is necessary. If T cell subpopulations 
are eliminated from the mouse, there is no tumor regression. The animals that are 
cured remain without tumor for 60 days and will resist a tumor rechallenge; therefore, 
they are truly immune. 
Dr. Rosenberg showed data from animal experiments in which metastatic disease in the 
lung had been established. The animals then received either unmodified or modified 
tumor cells and were treated with draining lymph node cells and IL-2 as in the human 
protocol. The impact of the treatment was determined. It was found that when TNF- 
modified tumor cells were administered, the number of metastases was decreased by 
Recombinant DNA Research, Volume 14 
[913] 
