Technical Abstract 
Scientific Abstract ' 
Our long term goal is to develop an effective intraperitoneal adoptive immunotherapy strategy for 
ovarian carcinoma utilizing TIL-(tumor infiltrating lymphocytes) derived T cells expanded in low 
concentrations of rIL-2. To achieve this goal it is important to determine whether ovarian TIL 
concentrate at metastatic tumor sites in vivo. Gene transfer technology offers a sensitive and 
reliable method to track TIL to tumor sites over a prolonged period of time. The specific objectives 
are to determine: (1) The feasibility of detecting CD3+CD8+ Tll^derived T cells that have been 
marked with the gene for Neomycin phosphotransferase encoded in a safety modified retroviral 
vector (GlNa) at tumor sites after IP injection, and (2) The fold of enrichment of CD3+CD8+ marked 
TIL at tumor sites. 
Background and Rationale 
Epithelial ovarian cancer primarily involves the peritoneal and serosal surfaces, a fact that has 
made this tumor a target for intraperitoneal (IP) biologic therapies including adoptive cellular 
therapy (1.2). Preclinical studies from our laboratory suggest that ovarian TIL may represent an 
acUve Immune response of the host directed against the tumor (3.4). T-cell lines from ovarian TIL 
expanded in low dose rIL-2 exhibited cytotoxicity primarily against autologous ovarian tumor cells. 
Our studies have shown that the TCR CD3 complex and CD8 antigens on rIL-2 expanded TIL, and 
MHC Class I surface antigens on ovarian carcinoma cells are involved in this in vitro cytotoxic 
activity (3.4). Ovarian TIL-derived T-cell lines have also been shown to produce substantial 
quantities of cytokines including interferon-y, TNF-a. and GMCSF following culture in low 
concentrations of rIL-2 (20 lU) (5). Production of these cytokines by CD8+ T-cell lines can be 
significantly enhanced by stimulation with autologous tumor cells. TIL may have therapeutic effect 
in vivo either by direct cell to cell contact or through the production of the cytokines described 
above. We have completed a pilot clinical protocol in which 8 patients with advanced platinum 
refractory ovarian carcinoma received IP rIL-2 expanded TIL and low dose rIL-2. Four of these 
patients had clinical activity after receiving IP TIL plus low dose rIL-2. In a preliminary study 
utilizing i^Un-labeled rIL-2 expanded TIL derived T cells, we observed an increase in the uptake of 
radioactivity in liver metastases of a patient with ovarian cancer. Neo^ marked TIL expanded in rlL- 
2 can be detected in both man and in animals for a considerable time after injection (6,10). 
Previous studies in patients with melanoma have shown that TIL from these patients were 
successfully transduced with the neo gene using the LNL6 retroviral vector (5) which has the same 
backbone as GlNa, the vector proposed for our marker protocol. The neo^ transduced gene cells 
could be Identified in biopsy material from melanoma metastases even at 64 days after injection. 
The hypothesis to be tested in this protocol is that ovarian rIL-2 expanded purified CD8+ TIL- 
derived T cells concentrate in metastatic tumor sites in vivo after IV injection. 
The experimental approach is briefly described as follows: Marking of the CD8+ TIL-derived T 
cells by infection with GlNa will be performed after positive selection has been accomplished, and 
after the purified CD8+ T cells are observed to be multiplying in a T-flask. At this point consenting 
patients will be registered. A mixture of bulk expanded transduced and nontransduced T cells will 
be injected IP wi^ rIL-2. Following treatment with the adoptively transferred TIL, (containing 
transduced and nontransduced cells) and IP rIL-2, biopsies will be obtained at laparoscopy 
performed at 1 month after injection of the TIL. from previously documented tumor sites and from 
normal tissues (peritoneum, omentum (if present) and lymph node tissue). Although animal and 
human studies show that injected TIL can be recovered beyond 2 months, there is evidence that 
selection of an earlier time point will maximize neo^ detection, with the requirement of a smaller 
number of patients. DNA will be extracted by a proteinase K phenol procedure and the neo^^ 
specific DNA quantitated by PGR according to Morgan et al (12). A successful result is defined 
when the mean number of marked cells in tumor tissue is significantly larger than the mean 
number of marked cells in blood or normal tissues. The study will be completed with a maximum 
of 9 and a minimum of 5 evaluable patients to determine the fold increase at metastatic tumor 
sites. The proportions of CD3‘'’CD8''' T cells will be determined in biopsies utilizing OKT8 and OKT3 
moAbs by histochemical procedures that we have established. In addition to these objectives, we 
will also determine the persistence of neo^ marked T cells in the abdominal cavity during the course 
of treatment with TIL + rIL-2. 
Recombinant DNA Research, Volume 19 
[453] 
