1546

Pancreatic cancer, one of the most insidious forms of cancer, is the fourth leading cause of all cancer deaths in both women and men. Due to its retroperitoneal location and silent growth, most patients present when it is too late to undertake curative treatment. This translates to a 5-year survival of less than 3%. Accordingly, we have generated MAb-PAM4, an anti-MUC1 antibody, and developed an immunoassay to quantitate PAM4-reactive MUC1 antigen in the sera of cancer patients. Although there have been several reports concerning anti-MUC1 MAbs reactive with several types of cancer, MAb-PAM4 identifies an epitope that is more restricted to pancreatic cancer. The immunoassay consists of MAb-PAM4 as the capture reagent and an IgG fraction derived from a polyclonal, anti-MUC1 antiserum as the probe reagent. The former reagent provides specificity and the latter reagent high signal output. At a cutoff value of 10.2 units/ml, 41 of 53 (77%) pancreatic cancer patients were positive, with a median value of 31.7 units/ml. None of the 20 normal individuals and only 2 of 32 (6%) patients with pancreatitis were positive at this cutoff value. ROC analyses for discrimination of pancreatic cancer from normal and pancreatitis specimens provided an area under the curve (AUC) of 0.88 (95% CI, 0.81 - 0.95), with a specificity of 96% and a positive likelihood ratio of 20.5. Of the 53 pancreatic cancer specimens, 41 were assessable for both PAM4-MUC1 and CA19-9. Of these, 24 (59%) were considered positive for CA19-9 (at a cutoff of 35 units/ml and a median value of 99 units/ml). As with the PAM4-MUC1 assay, none of the normal specimens was positive. However, of the 32 pancreatitis samples, CA19-9 was positive in 38%. ROC analyses for discrimination of pancreatic cancer from normal and pancreatitis serum specimens provided an AUC of 0.74 (95% CI, 0.63 - 0.84), with a specificity of 77% and a positive likelihood ratio of 2.59. Statistical analyses for PAM4-MUC1 in this same subset of pancreatic cancer sera differed little from the group analyses discussed above. Sensitivity for this subset was slightly reduced (71%), but specificity remained high (96%), as did the positive likelihood ratio (18.7). No correlation was observed between PAM4-MUC1 and CA19-9 results. These initial data suggest that the PAM4-MUC1 assay may offer a significant improvement over the CA19-9 assay for the discrimination of pancreatic cancer from normal and pancreatitis populations. The results now need to be extended to larger numbers of specimens, including early pancreatic cancer, as well as various benign and malignant conditions, to determine the true clinical value of this new immunoassay. (Supported in part by grants CA92723 and CA98488 from the NIH.)

[Proc Amer Assoc Cancer Res, Volume 46, 2005]