Search

The cosecretion of insulin and islet amyloid polypeptide (IAPP) is altered when isolated rat pancreatic islets are incubated in culture media conditioned by human pancreatic cancer cells.Pancreatic cancer is usually associated with impaired glucose tolerance. This study investigates the tumor-derived influence on beta-cell secretion of pancreatic islets.Four conditioned media were prepared from two human pancreatic cancer cell lines (Panc-1 and HPAF), a hamster pancreatic cancer cell line (PC-1), and a fibroblast cell line (Ag1523). Isolated rat pancreatic islets were incubated first in the conditioned media or nonconditioned control medium for 24 h, then in the same kind of media containing 100 microM carbamylcholine for 90 min. Insulin and IAPP secretion were measured by radioimmunoassay.Islets in media conditioned by Panc-1 and HPAF cells demonstrated dissociation of insulin and IAPP secretion. During 24-h incubation, the dissociation was expressed as selectively decreased insulin secretion. With addition of 100 microM carbamylcholine, the dissociation was expressed as normal secretion of insulin and hypersecretion of IAPP. As a result, the IAPP/insulin molar ratios were increased in both groups during both time periods. The islets in PC-1 and Ag1523 media did not show any significant changes in insulin and IAPP secretion.

Alterations of glucose metabolism occur frequently in association with pancreatic cancer. Whether diabetes predisposes or is a consequence of pancreatic cancer has not been settled. We have found that induced diabetes prevents development of pancreatic cancer in the hamster model and that genetically diabetic Chinese hamsters resist pancreatic cancer induction, whereas the nondiabetic strain does not. Also, preliminary clinical data indicates that diabetes is associated with pancreatic cancer and that the two conditions develop simultaneously. We have found diabetes or impaired glucose tolerance in 76% of the pancreatic cancer patients at the time of cancer diagnosis. The diabetic condition improved after 75% pancreatectomy. Although the preliminary observations suggest that diabetes accompanies the disorder of pancreatic cancer, additional data are needed before any conclusion can be drawn.

Our studies have shown that almost all preneoplastic and neoplastic tumors induced in hamsters contain endocrine cells. Because the induced pancreatic cancer resembles the human disease in many respects, it was of interest to find out whether human pancreatic cancer also produce endocrine cells. To examine the participation of the endocrine cells in the exocrine pancreatic cancer, we first examined 32 human pancreatic cancers with the antibodies against insulin, glucagon, somatostatin, PP, VIP, gastrin, and calcitonin or by argyrophilic reaction (1). Thirty (79%) cancers showed ei therfew or many ceils immunoreactive with all but anti-gastrin-, -VIP, and -calcitoniu. Many cells showed an argyrophilic reaction. In 17 cases, the number of stained cells was excessive and the morphology of tumors was consistent with mixed ductal-islet tumor. Argyrophilic ceils and cells immunoreactive with antiinsulin and -glucagon were the predominant in such cases. However, becat~se the biopsy material we used was small and only a few sections from each tumor were examined, it was not possible to determine whether several types of immunoreactive cells reside at the same location, or whether some cells produce more than one peptide. In subsequent studies, we examined larger sections of surgical material from 9 pancreatic cancer patients by serial sections and by using a multilabeled immunohistochemical technique developed at our institute. With this method, we visualized 4 peptides in the same islet. In this study, we also included antibodies against other pancreatic hormones, including pancreastatin, chromogranin A, serotonin, and neuron-specific enolase (NSE). We found that all 9 cancers contained cells immunoreactive with all of the antibodies used, except for pancreastatin, which was missing in 3 cases ,and for PP, which was not detectable in two cases, Here again, insulin was the predominant cell type, followed by glucagon and somatostatin ceils. Many of these hor-mones were found within the same glands. Colocalization of some hormones, such as insulin and serotonin and NSE and glucagon, was indicated. Similar to the situation in hamsters, most endocrine cells were localized on the base of the glands, but many were also within the epithelium and some seem to reach the lumen by a slender neck. In several cases, accumulation of endocrine cells was found in the papillary projections of the malignant epithelium. Because endocrine cells are found at different levels of the malignant epithelium and within the lumen, it seemed that the endocrine ceils migrated within the epithelium, were shed into the lumen like the tumor cells and were replaced by new endocrine cells at the base of the malignant glands. Although many of these endocrine cells resembled the corresponding endocrine cells in the normal islets, some of these endocrine cells were atypical. Some had granules on the supranuclear cell portion. A few endocrine cells were found in the invasive portion of the cancers, including those that had invaded the periuneural space. There were also atypical cells with cytoplasmic vacnolization and irregular granular distribution, possibly as a reflection of the altered synthesis, storage, or release of the granules. On the other hand, differences in the staining intensity could indicate their endocrine activity. Consequently, it appears that in both humans and hamsters, endocrine cells participate in the development of exocrine pancreatic cancer, and these endocrine cells seem to be malignant and have endocrine activity. Nevertheless, participation of endocrine cells in the exocrine pancreatic cancer is conceptually and practically important, particularly in view of epidemiological studies, which link pancreatic cancer with diabetes. In fact, it has been established that more than 80% of pancreatic cancer patients develop frank diabetes or altered glucose tolerance. Our immunohistochemical and experimental studies, as well as clinical studies by Permert and associates (2-4), have led us to believe that endocrine cell proliferation in exocrine pancreatic cancer may cause alteration in hormonal homeostasis leading to glucose metabolic abnormalities. However, studies are required to substantiate this view.