UAB 9405 
March 16, 1994 
PAGE - 3 
More pertinent to this proposal, similar observations have been made in colon 
cancer patients (8,9,10). Hollingshead (8) reported that low tumor burden colon 
cancer patients immunized with a partially purified “tumor-associated antigen” 
preparation derived from allogeneic colon cancer sonicated membrane extracts 
mixed with complete Freund adjuvant and administered at multiple sites monthly 
times 3 could induce an immune response as determined by delayed 
hypersensitivity skin tests and migration inhibition assays to the “tumor- 
associated antigens.” Hoover and Hanna (9,1 1,12) utilized a similar group of low 
tumor burden (Dukes’ B and C) colon cancer patients who received autologous, 
enzyme disassociated, live, irradiated tumor cells (10^) mixed with BCG weekly 
times 2 and tumor cells alone week 3. This randomized trial provided evidence 
for delayed hypersensitivity skin tests (12) to autologous tumor cells (versus 
normal mucosal cells) and a reduced relapse rate (9,11). A very interesting and 
provocative recent report by the Memorial Sloan-Kettering group (10) utilized 
vaccine composed of a partially desialylated bovine submaxillary gland mucin 
alone (n=6) with DETOX (n=8) or with BCG (n=6) in colorectal cancer patients 
with low tumor burden (Dukes’ B, C and D with no evidence of tumor). This 
preparation contains TN and sTN antigens expressed on various adenocarcinomas. 
The sTN is the epitope for B72.3 and CC49 monoclonal antibodies. This trial 
reported human antibody responses (IgM and IgG) in patients receiving the mucin 
plus adjuvant and not in the mucin alone group. All patients received pre-therapy 
with cytoxan. Only 1 patient developed a positive delayed hypersensitivity skin 
test. 
There are numerous other trials with these and other tumor types which provide 
similar observations. Overall, it seems reasonable to say that some patients 
receiving active specific therapy with tumor vaccines have had evidence of 
immune responses and occasional tumor regressions or other evidence of efficacy. 
However, the antigens involved are ill defined, measures of immune response are 
primarily skin tests and some antibody response data. Methodology to evaluate T 
cell responses which are so prominent in animal model systems of active specific 
immunotherapy (13,14) is seldom included or not possible. 
2.2 CEA represents a reasonable tumor relevant antigen for a tumor vaccine. 
Carcinoembryonic antigen (CEA) is an oncofetal antigen present predominantly 
in fetal gutand adenocarcinomas of the colon, breast, lung, etc. It has also been 
identified in small amounts in normal adult colonic mucosa. CEA is a 180Kd 
glycoprotein expressed on the cell surface and is probably the most extensively 
characterized tumor-associated antigen in man (15,16). The CEA gene family 
belongs to the immunoglobulin super gene family and resides on the long arm of 
chromosome- 19. Other members of this family which share significant homology 
with CEA (e.g., non-specific cross-reacting antigen or NCA) are found in other 
normal tissues including gallbladder and leukocytes (16,17). A large number of 
monoclonal antibodies to CEA and other family members have allowed extensive 
epitope mapping which define CEA specific epitopes (antigenic sites) as well as 
cross-reactive epitopes and epitopes specific for other family members^ (18). 
Recombinant DNA Research, Volume 19 
[ 400 ] 
