Proteins in blood reveal pancreatic tumors.

Although patients diagnosed with pancreatic cancer rarely live more than a year, the disease progresses slowly. Pancreatic cancer can take up to two decades to become symptomatic, and for some of that time it’s potentially curable with surgery. Unfortunately, patients generally aren’t diagnosed with asymptomatic disease—which has spurred a long, frustrating effort to identify blood markers that reveal new pancreatic tumors. The most widely adopted blood marker for pancreatic cancer is carbohydrate antigen 19-9 (CA19-9), which clinicians measure to monitor cancer progression. But the protein’s levels in the blood also rise with other illnesses, including diabetes, chronic pancreatitis, and gastrointestinal cancers. Moreover, 5%–10% of humans can’t produce CA19-9. Even so, research on using CA19-9 to screen for pancreatic cancer continues. The American Association for Cancer Research held a special conference on pancreatic cancer in New Orleans on May 19. Findings reported there show that combining CA19-9 with three other proteins improves its screening resolution. Blood samples evaluated came from subjects with early-stage tumors, and “that definitely set our investigation apart from other research,” said Ayumo Taguchi, M.D., Ph.D., assistant professor at the University of Texas M. D. Anderson Cancer Center in Houston, who presented the results. “Most of the literature describes CA19-9 levels from patients with stage III or IV disease,” he said. The combined protein panel was better than CA19-9 alone at distinguishing people with pancreatic cancer from healthy control subjects, and from individuals with pancreatic cysts or chronic pancreatitis. Matthew Firpo, Ph.D., associate professor of surgery at the University of Utah School of Medicine in Salt Lake City, described the findings as an important step forward. Firpo and colleagues supplied 42 of the 98 cancer-related samples from a pretreatment collection at the University of Utah. Many samples came from people diagnosed incidentally with pancreatic cancer while being imaged for other abdominal problems, he said. But the study also had shortcomings. Although the panel in general performed better than CA19-9 alone, most reported differences were not statistically significant. The panel also wasn’t accurate enough to adequately minimize false-positive results.