1.0 BACKGROUND AND RATIONALE 
Cytotoxic T lymphocytes (CTL) recognize peptide antigens presented on the cell-surface in the 
context of major histocompatibility complex (MHC) class I molecules. Most CTL responses 
require T helper cell collaboration which provides for local secretion of cytokines such as 
interleukin-2 (IL-2). It has been hypothesized that the failure of most cancers to elicit an 
effective CTL response is, in part, due to a failure of T cell help. 
This hypothesis has been tested by providing a continuous supply of the helper cytokine IL-2 at 
the tumor site. In earlier studies, repeated local tumor injections of IL-2 were employed 1 , but 
with the availability of plasmid or retroviral-based expression vectors, tumor cells can be 
genetically-engineered to produced IL-2 2 ' 10 . The following principles have emerged from these 
studies: 1. poorly immunogenic tumors engineered to produce IL-2 have reduced tumorigenicity 
and can be rejected in vivo . 2, mice acquire protective immunity against parental cells, 3. the 
response is CD8-dependent, 4. the response is dose-dependent, 5. irradiated IL-2 producing 
tumor cells generated systemic immunity and 6. syngeneic , allogeneic, xenogeneic cytokine- 
producing cell lines, admixed with parental tumor effectively abrogates tumorigenicity and 
generates systemic immunity. 
These observations provide the experimental foundation for using genetically-engineered vaccines 
to generate cell-mediated immunity in human cancer patients. However, there are several 
logistical problems in generating cytokine gene-transduced autologous tumor cell lines for use 
as vaccines. The establishment of a tumor cell line for each patient is expensive and labor- 
intensive and requires the transduction, selection, characterization and biosafety testing for each 
patient. Some tumor types cannot be adapted to tissue culture. Moreover, tissue culture may 
result in antigen drift or selection in vitro of nonrepresentative tumor cell populations. 
Another approach is to exploit the bystander effect of a cytokine production well documented 
in animal models. An effective local and systemic immune response can be generated with 
unmodified tumor cells if they are admixed with cytokine-producing cells of any histological 
type. 
Our approach will be to use a well characterized melanoma cell line (M-24) engineered to 
produce human recombinant IL-2. The M-24 melanoma cell line has been safely used as an 
allogeneic melanoma vaccine in hundreds of melanoma patients . 11,12 It has also been used in an 
FDA and IRB - approved trial in melanoma patients receiving systemic IL-2 . (see ref 11 in 
Appendix II) Vaccines will consist of fresh (or cultured) autologous tumor cells admixed with 
IL-2 transduced M-24. This approach has several theoretical advantages: (1.) patients receive 
a well-characterized and biosafety-tested genetically-engineered cell line producing reproducible 
levels of IL-2, (2.) higher levels of cytokine production can be achieved since higher producer 
clones are isolated, (3.) allogeneic cell lines may incite a strong local inflammatory response 
above that generated by IL-2 production, and (4.) if effective, larger numbers of patients with 
different tumor types could be treated. 
[420] 
Recombinant DNA Research, Volume 18 
