However, these cells, when transduced with IFN-y generated a CD8+ T cell immune response that was against 
I wild-type tumor cells.l3 The relationship between MHC class-I expression and tumor rejection may be complex, as 
higher levels of IFN-y secretion are required in murine adenocarcinoma cells to elicit tumor rejection than are required 
to induce MHC class-I antigens.l^ Therefore, induction of tumor specific immunity by transduction of IFN-y may 
! include enhanced MHC class-I expression and other variables. The fact that these experiments have been successful 
! in poorly immunogenic, syngeneic tumors of nonviral origin suggests that this strategy may be successful for 
: human tumors of relatively low immunogenicity. 
I 
Tumor infiltrating lymphocytes (TIL) derived from patients whose melanoma cell lines had normal expression of 
HLA-A2 antigens were CD8-1- and were capable of lysing autologous melanoma cells. 15 In contrast, TIL derived 
from patients whose melanoma cell lines did not express HLA-A2 were predominantly CD4-1- and were not 
cytotoxic. These studies support the notion that enhanced MHC class-I expression could result in greater specific 
immunologic responses to tumors. Transduction of melanoma cells with IFN-y enhanced ejmression of HLA 
molecules and their ability to stimulate autologous human uimor infiltrating lymphocytes (TIL).lo 
Treatment of human neuroblastoma cells with exogenous IFN-y can increase expression of HLA class I antigens and 
6-2 microglobulin, both in vitro and in patients. 1^-1 9 Also, IFN-y treatment of neuroblastoma cells in vitro 
increases their susceptibility to lysis by LAK cells and monocyies.20.2l 
The relationship between N-myc expression and MHC class-I expression in neuroblastoma has been investigated. In 
33 neuroblastomas, amplification of N-myc appeared to correlate with low levels of MHC class-I expression.22 
Fusion of N-myc overexpressing human neuroblastoma cells with cells not expressing N-myc resulted in a decline in 
N-myc expression and expression of class-I MHC genes.23 The apparent correlation between N-myc expression and 
expression of MHC class-I antigens is not absolute, however, and in a survey of 24 human neuroblastoma cell lines, 
N-myc expression was not invariably associated with low levels of 62 microglobulin expression.24 in studies of the 
MHC class-I promoter, distinct elements susceptible to N-myc mediated supression have been identified, and N-myc 
expression has been correlated with reduced binding of a transcription factor found to activate the MHC class-I 
enhancer.25 This data, although inconclusive, suggests a relationship between N-myc expression and MHC class-I 
down regulation. We have introduced IFN-y vectors stably into several patient derived neuroblastoma cell lines. 
Levels of IFN-y production have ranged from 12 to 1550 u/ml and have persisted for up to 2 months in vitro. 
Transduction into the low MHC-I expressing LAN-6 cell line has increased MHC-I expression up to 1000 fold as 
measured by FACS analysis. Less striking inaeases in MHC-II expression have also been noted. In addition, 
upregulation of the neuroblastoma tumor associated antigen HSAN 1.2was noted in transduced SKN-AS cells. This 
suggests that IFN-y will be highly useful in enhancing immunogenicity of neuroblastoma cells in vivo. 
2.2 Adult Studies. There have been no publications dealing with immunization utilizing EFN-y transduced 
human uimor cells. One protocol for patients with melanoma has been approved by the RAC and is open at Duke 
University. This protocol employs the same retrovirus for transduction of IFN-y cDNA that we shall utilize 
(provided by Viagene, Inc.). 
23 Pharmacology. Pharmacokinetics and biological activity in subcutaneous long-term administration of 
recombinant BFN-y in cancer patients have been investigated.^ In doses of 0.1, 0.25, or 0.5 mg/m^ in two or three 
injections per week for up to 180 days there was limited intra-individual variability with respect to the plasma 
concentration/time profiles. However, the interindividual variation was high for the parameters area under the data 
points and maximum plasma concentration. Despite this high interindividual variability, the mean values obtained 
for were approximately proportional to the dose administered (0.1, 0.25 or 0.5 mg/m2). In this study, objective 
antitumour responses were observed in two patients, but there was no partial or complete remission. 
Summary. We aim to maximize anti-tumor immune responses to neuroblastoma cells by increasing their 
immunogenicity through genetic engineering. Our large bank of neuroblastoma cell lines and matched normal 
leukocytes provides a unique opportunity to determine if partially MHC matched tumor cells can be effective 
immunogens. The availability of patients with measurable disease as well as no detectable disease after ABMT 
provides two distinct settings for testing the toxicity and immunogenicity of genetically engineered autologous and 
allogeneic neuroblastoma cells. The proposed studies are relevant to a variety of other high risk human tumors for 
which intensive, cytoreductive therapy followed by immunologic therapy may be an effective treatment 
Recombinant DNA Research, Volume 19 
[99] 
