Summary The expression of a novel tumour associated antigen CA 242, defined by the monoclonal antibody C 242, was studied by immunoperoxidase staining in formalin-fixed, paraffin-embedded tissue sections from normal pancreata, pancreata with pancreatitis and benign and malignant pancreatic neoplasms. The antigenic determinant of the C 242 antibody is a sialylated carbohydrate structure, related but chemically different from tumour marker antigens CA 19-9 and CA 50. Thirty-eight of 41 (93%) well to moderately differentiated ductal adenocarcinomas of the pancreas and all cystadenocarcinomas were positive for CA 242. , corresponding to sialylated blood group antigen Lewisa, which is the antigenic determinant detected also by the 19-9 antibody (Koprowski et al., 1979;Magnani et al., 1982). In addition, the C 50 antibody reacts with at least one other carbohydrate structure, sialosyllactotetraose .It has been proposed that CA 242 appears in serum on the same macromolecule as CA 50 and CA 19-9 (Lindholm et al., 1985). Recently a DELFIA assay for quantitation of CA 242 in serum has been described (Nilsson et al., 1988 cinomas (38 primary tumours and 10 metastatic tumours); five anaplastic carcinomas, seven cystadenomas, three cystadenocarcinomas and nine neoplasms of endocrine origin. The samples were formalin-fixed, paraffin-embedded surgical specimens, which had been stored for between 6 months and 10 years.
AntibodiesTissue culture supernatant containing mouse monoclonal antibody C 242 (IgGi) was used for the CA 242 stainings. The optimal dilutions of primary antibodies were determined in control series.Staining procedure Paraffin sections 5 jm thick were deparaffinised and treated with 0.4% pepsin (2,500 FIP-U g-'; Merck, Darmstadt, FRG) in 0.01 N HCI for 1 h at 37°C. All sections were then incubated in 0.5% hydrogen peroxide in methanol for 30 min to block endogenous peroxidase, incubated with non-immune horse serum, diluted 1: 20, and then reacted with the C 242 antibody, diluted 1: 50. Bound antibody was visualised by the avidin-biotin complex assay (ABC; Vectastain, Vector, Burlingame, CA). The sections were successively treated with biotinylated anti-mouse immunoglobulin antiserum, avidin and biotinylated horseradish peroxidase complex. Each step was followed by washing in phosphate-buffered saline (PBS). Finally, sections were incubated with 3-amino-9-ethylcarbazole (AEC) and hydrogen peroxide, and then counterstained with haematoxylin. Sections stained with nonimmune mouse serum and PBS, respectively, instead of the primary antibody served as negative controls.The effect of enzyme pretreatment was tested in a series where CA 242 positive sections were pretreated, with 0.4% pepsin in 0.01 N HCI, with 0.01 N HCI only or with PBS. The optimal staining reaction was obtained after pepsin treatment for 1 h. None of the negative specimens became positive after pepsin treatment.An arbitrary scoring of distribution using + or + + was used. Specimens including both carcinoma and normal pancreatic tissue were used f...