The failure to detect pancreatic cancer until after it has progressed to a fatal stage has long been the bane of doctors and the demise of patients. Imaging the out-of-the-way organ with MRI or CT scans often fails to catch tumors and, so far, there are no reliable blood-based markers for the disease.

For most patients--including actor Patrick Swayze, who died this week as a result of this aggressive malignancy--abdominal pain, weight loss and jaundice are the disease's first indicators, and their onset usually occurs after the cancer has metastasized, erasing any possible benefit of surgery.

Approximately 42,000 people in the U.S. will be diagnosed with pancreatic cancer this year. Nearly all of them will succumb to the disease, 76 percent within the first year after diagnosis. Swayze managed to battle his cancer for 20 months, 12 more than the median survival time. Because pancreatic tumors are often resistant to chemotherapy or radiotherapy, "the only hope is that we can remove it surgically" at an early stage, says Mark Bloomston, an associate professor in the Division of Surgical Oncology at The Ohio State University Medical Center in Columbus.

New hope for improved detection comes in the form of small RNA molecules called microRNAs. Unlike the better-known role of RNA as an intermediary between genes and proteins, these shortened versions of 19 to 25 nucleotides do not get translated into proteins. Instead, a microRNA can bind to a target RNA strand with a complementary sequence, similar to how two strands of "matched" DNA pair up, and prevent that RNA from getting made into a protein. If the targeted RNA encodes a tumor suppressor protein, as scientists have found for some, the ultimate effect of the microRNA could be to promote cell--and tumor--growth.

Recent studies have deciphered a microRNA "signature" associated with cancer: The levels of certain microRNAs in tissue from different cancers are abnormal--usually more abundant--compared with the tissues from healthy people. For pancreatic cancer, scientists have catalogued dozens of microRNAs whose levels are different than in healthy samples.

To spin this connection into a new tool for cancer detection, a team of researchers at the University of Texas M. D. Anderson Cancer Center in Houston looked for differences in microRNA levels in a more accessible type of patient sample--blood. Out of the dozens of choices, they picked four microRNAs to measure. "This was a proof of principle; we wanted first of all to see whether or not we could detect any of these microRNAs that had been associated with pancreatic cancer in the plasma," says Subrata Sen, a professor of molecular pathology at M. D. Anderson. Not only did the group detect these microRNAs circulating in the blood, they found their levels were higher in the blood of pancreatic cancer patients compared with healthy control subjects. Their results were published last week in Cancer Prevention Research.

Because these four microRNAs are also more abundant in other cancers, such as those of the lung and breast, this quartet cannot be used to identify the type of cancer. "We are working right now on finding tissue-specific microRNAs" that could have the potential to screen specifically for pancreatic cancer, Sen says. Two of the four the group chose are associated with different stages of malignancies, so doctors could combine them with the eventual, pancreas-specific microRNAs to get a set that also diagnoses the stage of pancreatic cancer.

Even if measuring microRNAs does not turn out to be sensitive enough for general screening, it would arm doctors with a great tool for screening at-risk individuals, Bloomston says. Of course, singling out those groups is half the battle in a disease such as pancreatic cancer. Those with a family history, which accounts for about 10 percent of pancreatic cancers, would make up one obvious group. Sen and his colleagues are currently studying gene mutations that could play a part in pancreatic malignancies, and also provide a preliminary tool to identify more susceptible individuals. Researchers have also linked drinking and smoking to this form of cancer, so people who partake in these activities might also qualify for microRNA screening.

Pancreatic cancer is the third type of cancer, after prostate and ovarian, to be associated with altered levels of microRNAs in the blood. A blood test for microRNA levels could screen patients at risk for more than one type of cancer or, as scientists gain better knowledge of tissue-specific microRNAs, specific cancers. And, one day, screening may not even involve a needle if, as Bloomston points out, scientists can confirm levels in urine and saliva.