[Frontiers in Bioscience 3, d1148-1160, November 15, 1998]

Table of Conents
 Previous Section   Next Section


Shamlal Mangray and Thomas C King

Department of Pathology and Laboratory Medicine, Brown University School of Medicine, Providence, RI

Received 8/24/98 Accepted 9/4/98


The ideal screening test for pancreatic carcinoma would be a specific marker detectable in peripheral blood or feces which would enable noninvasive screening and detection of early tumors prior to the onset of symptoms. No such test is currently available and the prospects for developing such a marker are not promising. K-ras mutations have been assessed in stool samples and this marker has shown good sensitivity but limited specificity for pancreatic adenocarcinoma (68). Positivity can also result from non-neoplastic conditions of the pancreas such as pancreatitis, as well as from adenomatous colon polyps.

Recently Nomoto et al advocated the clinical application of mutations of the K-ras oncogene for the detection of micrometastases (69). Using a hemi-nested polymerase chain reaction (PCR) and restriction fragment length polymorphism (RFLP) analysis to detect mutations of K-ras at codon 12, tumor cells could be detected in liver and lymph node tissue that were not apparent on histological examination. The authors suggest that this technique can detect subclinical micrometastases. This assay detected tumor derived K-ras mutations in the peripheral blood of all pancreatic cancer patients tested at the time of surgery emphasizing the degree of systemic involvement in this disease. A related PCR strategy has been successfully employed for the detection of tyrosinase mRNA in the blood or lymph nodes of patients with malignant melanoma and carcinoembryonic antigen mRNA in lymph nodes from patients with colon cancer(70). Early results in these systems suggest that molecular detection of tumor cells can be predictive of adverse outcome.

p53 mutations can be detected in duodenal brushing specimens and their presence has been shown to correlate with the presence of pancreatic adenocarcinoma (71). Iwao et al also suggested that detection of a specific p53 mutation at codon 273 (GGT to CAT in exon 8) was predictive of massive metastases and may be an indicator of unusually poor prognosis. This finding is provocative but must clearly be reproduced by other investigators. Caution must be used in basing a diagnosis of pancreatic adenocarcinoma on the presence of p53 mutations alone, since p53 mutations have been identified in dysplastic and precursor lesions in other systems.

Rozenblum et al pointed out that the coexistence of K-ras and p16INK4a alterations is quite uncommon in tumors other than pancreatic adenocarcinoma in which they found 83% concordance (21). They further suggested that this concordance could be used as a diagnostic criterion for pancreatic adenocarcinoma. This observation is interesting scientifically but has no obvious application to clinical diagnosis or management of patients with pancreatic adenocarcinoma.

Recently, anti-p53-autoantibodies (a-p53-aab) have been investigated as a potential tumor marker for pancreatic adenocarcinoma. a-p53-aab can develop as a humoral response to increased intracellular p53 protein in tumor tissue. Since p53 is frequently mutated and overexpressed in pancreatic adenocarcinoma, Gansauge et al. evaluated 145 patients with pancreatic adenocarcinoma for the presence of a-p53-aab using ELISA and Western-blotting (72). 16% of the pancreatic adenocarcinoma patients had demonstrable a-p53-aab in their serum. The presence of a-p53-aab was more frequent in higher stage patients with lymph node metastases Using similar methodology, Raedle et al. found a similar frequency of positivity for a-p53-aab in patients with pancreatic cancer (73). Unfortunately, a similar percentage of patients with chronic pancreatitis were also positive (12%) as well a lesser percentage of patients with acute pancreatitis (5%). a-p53-aab would not expected to be specific for pancreatic adenocarcinoma since p53 is overexpressed in a wide variety of tumors.

At the present stage of our knowledge of pancreatic carcinoma, none of the available screening methodologies are appropriate for the general population based on their cost and lack of specificity. Patients at high risk for the development of pancreatic adenocarcinoma based on genetic syndromes or strong familial clustering might benefit from some of these strategies, however. Evaluation of peripheral blood and/or liver and lymph node tissue from patients with pancreatic adenocarcinoma may be useful in identifying patients at high risk for local progression. On the other hand, overall prognosis for pancreatic adenocarcinoma is currently so poor that such information may not be of great clinical utility. With the development of more effective therapeutic strategies for pancreatic adenocarcinoma (see below), this type of prognostic information may be of considerable value in the future.