lumorigenicity and has induced persistent anti-tumor immunity in several animal models. In a renal cell cancer 
model, extent of modulation of MHC class 1 and class II correlate with response to interferon gene therapy (30). 
In contrast to many other solid tumors, neuroblastoma cells grow readily in vitro (64,65). Cell lines are readily 
derived both from biopsy tissues as well as from bone marrow and occasionally peripheral blood. Ability of the 
human neuroblastoma to form a cell line is associated with poor prognosis (Figure 12). Thus neuroblastoma is an 
excellent candidate for gene therapy. The advantages of neuroblastoma as a model system include ability to grow 
cells in viux), ability to achieve complete remission with standard therapy, and the high and predictable relapse rate. 
Hence we wish to test modem gene therapy strategies for neuroblastoma using a retroviral vector designed to express 
y-IFN. The study is designed to address s^ety, clinical response (disease free interval, relapse rate, tumor burden) and 
biologic response (immune response). 
2.0 INTRODUCTION 
2.1 CANCER IMMUNOLOGY 
Host immunity is an important mechanism by which malignant tumors can be kept in check or even cleared. 
Established tumors are therefore under selective pressure to develq) means of evading the immune system. Although 
the mechanisms by which tumors evade the immune system are undoubtedly varied and complex, at least one 
evasion mechanism appears to derive from the fact that the levels of Class I MHC proteins are unusually low on 
some tumors (1,2). Since Class I MHC proteins are essential for the presentation of intracellular antigen to the 
immune system (3,4,66), suboptimal levels of MHC proteins may prevent effective presentation of tumor-specific 
antigens. The immune system is therefore relatively blind to the tumor cell, at least with regard to MHC Class I- 
restricted cytotoxic T-lymphocytes (CTL). A second mechanism by which tumors could evade the immune system 
is by developing anergy due to generation of suppressor T cells or other immune depressing substances to prevent 
the normal activation of anti-tumor immunity (immune "anergy")(5-7). This may include a block to MHC-II 
restricted help needed for generation of CTL. Effective immunotherapy strategy must necessarily overcome one or 
both of these phenomena. 
Methods for improving the immune responses against tumors, especially CTL responses, are likely to be 
therapeutically beneficial to cancer patients. A variety of approaches have been used to engender non-specific 
augmentation of the immune systems of cancer patients. For instance, lymphokines such as EL-2, a-EFN, and more 
recently, y-IFN have proven to be therapeutically useful, although efficacy is limited and toxicity associated with 
systemic administration is considerable (8-10, 73). y-IFN and EL-2 presumably function as growth and activation 
factors (11-16) to tumor-reactive immune cells such as CTL, macrophages, natural killer (NK), and lymphokine 
activated killer / tumor infiltrating lymphocytes (LAK/TEL) at the site of the tumor, possibly to overcome immune 
suppression. ylFN also increases the expression of MHC proteins on the surface of tumor cells to improve 
antigenic presentation (17,18,28,75,76). However, systemic administration of lymphokine may only marginally 
contribute to tumor rejection, and is often limited by significant toxicity. Systemic injection of lymphokines is 
thus a highly inefficient way to deliver these critically important proteins to their desired site of action, the tumor. 
One approach that would assure the continued presence of y-IFN locally around tumor cells is to introduce 
constitutively expressed lymphokine genes into tumor cells. The resultant constitutive, low level production of y- 
IFN should specifically increase the level of MHC (the antigen presenting proteins) on the surface of tumor cells. In 
addition, y-E^ should be secreted locally around the tumor and thus aid the growth and activation of local immune 
cells (such as CTL, macrophages, and NK cells, as they respond specifically to tumor antigen. The concentrated, 
local presence of the immune-activating y-IFN may thus help overcome immune anergy. The total amount of 
lymphokine should then be quite low, but it should be concentrated where it would be most effective, namely at the 
site of tumor cells. The ove^l effect of lymphokine gene transfer would be to "flag" the tumor cell as something 
that the immune system should carefully scrutinize and respond against. The resultant boost in tumor specific 
immunity could result in improved responses compared with non-gene-modified approaches (Fig 1). 
We (see section 3.2) and others (24-30) have observed the potential utility of retrovirally mediated lymphokine gene 
delivery in general, and y-IFN gene delivery specifically (28-30), in murine tumor model systems. TTie data have 
been sufficiently encouraging to warrant clinical investigation in cancer patients. We therefore propose to 
investigate the utility of this approach in human neuroblastoma. 
Recombinant DNA Research, Volume 19 
[91] 
