Where possible, lymphocytes from the draining lymph nodes, tumor, the vaccine 
injection sites, and peripheral blood will be examined for reactivity against 
a panel of HLA-A2 melanoma cell lines that do not share the allogeneic 
determinants of the immunizing cell line. Controls will include malignant 
non-melanoma HLA-A2 tumor cell lines and autologous EBV-trans formed 
lymphoblastoid cell lines (EBV-LCL) . 
Immunization with the B7- trans f ected melanoma cell line may also induce a CD4 + 
T cell response. Since class II MHC generally presents phagocytosed and not 
cellular antigens, induction of a CD4+ T cell response would likely occur by 
macrophage processing of dead tumor cells. It is interesting to note that CD4+ 
antigen-speci f ic T cell responses have been sufficient to induce tumor 
regression in some animal models, without the participation of CD8+ cells. 
It will be of some interest to determine whether our patients develop 
melanoma-specific CD4 T cell proliferative responses. 
We recognize several potential problems. Immune tolerance to the melanoma 
antigen (perhaps induced by recognition of antigen on tumor cells without 
costimulatory signals) may be present in vivo and may not be overcome by 
immunization with the B7- trans f ected cell lines. Shu et al have shown that 
immunization with tumor cells plus C. Parvum into the footpad of a non-tumor 
bearing mouse results in the development of tumor antigen-specific pre- 
effector CTL in the draining lymph nodes. Full maturation to CTL is 
accomplished by further culture of pre-effector lymphocytes in vitro with the 
sensitizing tumor and IL-2. However, in the presence of pre-existing visceral 
tumor (but not tumor in the subcutaneous tissue), immunization into the 
footpad fails to elicit the tumor-specific pre-effector T cells in the 
draining lymph nodes (23) . In another series of experiments (conducted at the 
BRMP) , Ochoa and his colleagues have demonstrated that T cells from animals 
with advanced tumors have defective T cell receptors and receptor signalling 
(24). Other mechanisms of tolerance have been described including T cell 
exhaustion by chronic exposure to antigen or the presence of a suppressor CD4 + 
lymphocyte population. Thus, it may be difficult to induce an antigen- 
specific CTL response in a patient with advanced melanoma by in vivo 
immunization, even though the immunizing cell line has been modified to 
contain costimulatory signals. 
Other problems related to the vaccine and the assay for response may be 
encountered. The effect of the allogeneic determinants present on the B7- 
transfected cell lines with regard to the immune response to the relevant 
antigen is unknown. We are also aware that the optimal frequency of 
immunizations has not been determined. The effect of restimulating activated 
Recombinant DNA Research, Volume 18 
[729] 
