Evaluation of a new tumor-associated antigen in pancreatic cancer.

This study was carried out to assess the diagnostic accuracy of a new gastrointestinal cancer antigen ( GICA ) defined by a monoclonal antibody. Its sensitivity and specificity were assayed in a group of patients with different pancreatic diseases (10 acute pancreatitis, 27 chronic pancreatitis, 22 cancers of the pancreas) and in 29 normal individuals. The concentration of GICA was always inferior to 37 units/ml (our discriminant limit between cancer and noncancer patients) both in cases with chronic pancreatitis and in healthy subjects. Increased levels of the antigen were found in 16/22 (72.7%) pancreatic cancer patients and in 3/10 (33.3%) cases with acute pancreatitis. The assay was within the normal range in 2 (28.6%) out of 7 cancers judged resectable. The test is simple and rapid, but its relative sensitivity and the frequent elevation of GICA in other adenocarcinomas of the gastrointestinal tract make it unsuitable for screening programs in pancreatic cancer. Even its use for early diagnosis of cancer of the pancreas does not seem promising. The major finding of our study is the lack of false-positives in patients with chronic pancreatitis and therefore the usefulness of this test in differentiating preoperatively between chronic inflammation and cancer of the gland. Frequent increase of the marker in patients with acute pancreatitis is not yet clear.