NEW ORLEANS — Pancreatic cancer kills swiftly and surely, and often goes undiagnosed until it's far too late for doctors to provide the only cure: surgery. But new research being reported here this week may give patients a better chance at early detection, firm diagnosis and, someday, better treatment options for the fourth leading cause of cancer death.

In 48 human pancreatic cancer tumors like this one, called an adenocarcinoma, CEACAM 1 was found to be elevated in all but two patients.

Today, researchers from the University of Michigan Comprehensive Cancer Center announced that they have found a protein which allows them to tell pancreatic cancer from normal tissue better than the current “gold standard” blood test used nationwide. They believe this protein, called CEACAM 1, could be used to detect early signs of cancer, especially in those patients at highest risk for the disease.

Yesterday, the same team reported new findings about the basic cellular processes that allow pancreatic cancer to develop. The discovery of “crosstalk” between two cell signaling pathways involved in the disease may yield clues as to why pancreatic tumors develop, and perhaps one day lead to innovative ways to stop them.

Also today, the U-M researchers reported initial results from gene-based efforts to tell pancreatic cancer from its more common imposter: chronic inflammation of the pancreas, or pancreatitis. By looking at expression levels for three genes in tiny tissue samples removed from patients through a slender needle, they were able to distinguish the two conditions. More research is needed, but the technique could lead to new ways of telling pancreatitis from cancer without the need for surgery.

All three announcements were made at the Digestive Diseases Week meeting here.

“We're definitely getting closer to the kinds of innovation that will be needed to increase patients' odds of finding and surviving pancreatic cancer,” says Diane Simeone, M.D., whose laboratory team made the cell signaling discovery and who co-leads the U-M effort to translate basic genetic research into clinically useful tests.

The newly announced findings all stem from research led by Simeone and Craig Logsdon, Ph.D. One year ago, they published a list of 158 genes that were over-expressed in patients with pancreatic cancer — the new findings draw on that list to focus in on proteins encoded by some of the most promising genes.

This graph shows that CEACAM 1 was being produced much more often in pancreatic tumor tissue than in normal tissue or inflamed tissue (pancreatitis). This evidence, combined with results from patients' blood tests, show that CEACAM 1 may be a good biomarker for pancreatic cancer.

The CEACAM 1 findings show that the protein is elevated in the blood of pancreatic cancer patients, and that levels of RNA and mRNA corresponding to it are far higher in the cells of cancer patients than in pancreatitis or normal tissue. CEACAM 1 was better at telling pancreatic cancer from normal tissue than the current “gold standard” test, a protein called CA19-9. A test combining the two proteins was even better.

Simeone, who runs the U-M CCC's multidisciplinary pancreatic cancer clinic and operates on dozens of patients a year, notes that CEACAM 1 may be an important biomarker for the early stages of pancreatic cancer, before it becomes a full-blown tumor. In 17 of 20 samples from patients in the final pre-cancerous stage of pancreatic cancer, CEACAM 1 was overexpressed.

In all, the U-M team performed Affymetrix RNA “gene chip” assays on 10 pancreatic cancer tissue samples, five pancreatitis samples and five normal samples. The levels of CEACAM 1 RNA were far higher in all cases. The same was true when 89 samples were examined using quantitative rapid polymerase chain reaction RNA analysis. And, the same was true in blood samples from several dozen patients.

CEACAM 1 was present even in pre-cancerous pancreas lesions like this one, meaning it may be useful for detecting cancer early in people at high risk for the disease, such as smokers and those whose families are prone to pancreatic cancer.

More research is needed on the ability of CEACAM 1 serum tests to tell pancreatic cancer from pancreatitis, Simeone says, and a larger validation trial of the biomarker will begin this summer via a multi-center network of investigators. She hopes that some day, patients may be able to have a blood test for a “panel” of proteins, especially if they are at high risk for the disease because of family history, age or smoking history. Tobacco use elevates the risk of pancreatic cancer greatly.

The need for improved diagnosis is great: 30,000 Americans each year die from pancreatic cancer, and 80 percent are diagnosed long after surgery to remove their tumor is possible. The symptoms of pancreatic cancer often mimic other diseases, or are overlooked until it is too late. Only 3 percent of all patients live even five years after diagnosis, and although pancreatic cancer is only the 10 th most common cancer, it is the fourth leading cause of cancer death.

Progress against this deadly disease will rely on basic research, in addition to clinical innovation. Even as the CEACAM 1 clinical research continues, U-M researchers are pursuing important clues to the processes inside the cell that govern the growth of pancreatic tumors.

In a plenary lecture at the DDW meeting, postdoctoral fellow Gangyong Li, Ph.D., of Simeone's team presented new evidence about an interaction between two important cell-signaling pathways — TGF Beta and protein kinase A (PKA). The presentation follows on a paper that the team published earlier this year, showing the existence of a previously unknown mechanism by which proteins spurred by TGF Beta can activate PKA.

Both pathways are crucial to the regulation of cell growth, a process that goes haywire when pancreatic cells become cancerous. At the DDW meeting, Li will show how two proteins whose action is governed by TGF Beta, called Smad3 and Smad4, interact with a subunit of the PKA molecule. This interaction, the team believes, spurs the growth of tumor cells in ways that hadn't previously been known.

“To better understand how things go wrong in cancer, you need a better understanding of the normal signaling process when things are going right,” says Simeone, an associate professor of surgery at the U-M Medical School. And eventually, she says, a better picture of what molecules and pathways are important to the initiation and growth of pancreatic cancer may yield new opportunities for treatments tailored to particular types of pancreatic tumors.

Meanwhile, research led by U-M gastroenterologist Michelle Anderson, presented in a poster at the DDW meeting, may help spare some patients surgery. By looking for gene expression levels of three genes in samples of pancreatic tissue taken by fine-needle aspiration biopsy, Anderson and her colleagues found that they could pinpoint the nature of previously indeterminate lesions. If the results can be replicated in a larger clinical trial currently being planned, this approach could help prevent surgery for patients whose biopsy shows they have pancreatitis, not cancer.

The research team receives funding from the Michigan Life Sciences Corridor, the Lustgarten Foundation for Pancreatic Cancer Research, the U-M Comprehensive Cancer Center , and the National Cancer Institute. In addition to Simeone, Logsdon, Li and Anderson, the research team includes Baoan Ji, Ph.D., who presented the CEACAM 1 data, Beth Weinman, James M. Scheiman, Lizhi Zhang, Vijaya Ramachandran, David Hu, Michael Uhler, Thiruvengadam Arumugam, Samir Hanash, Thomas Giordano, Joel K. Greenson and Jeremy M. Taylor.

Contact: Kara Gavin

E-mail this information to a friend

Recent Press Releases