Years before they show any other signs of disease, pancreatic cancer patients have very high levels of certain amino acids in their bloodstream, according to a new study from MIT, Dana-Farber Cancer Institute, and the Broad Institute.
This finding, which suggests that muscle tissue is broken down in the disease’s earliest stages, could offer new insights into developing early diagnostics for pancreatic cancer, which kills about 40,000 Americans every year and is usually not caught until it is too late to treat.
The study, which appears today in the journal Nature Medicine, is based on an analysis of blood samples from 1,500 people participating in long-term health studies. The researchers compared samples from people who were eventually diagnosed with pancreatic cancer and samples from those who were not. The findings were dramatic: People with a surge in amino acids known as branched chain amino acids were far more likely to be diagnosed with pancreatic cancer within one to 10 years.
“Pancreatic cancer, even at its very earliest stages, causes breakdown of body protein and deregulated metabolism. What that means for the tumor, and what that means for the health of the patient — those are long-term questions still to be answered,” says Matthew Vander Heiden, an associate professor of biology, a member of MIT’s Koch Institute for Integrative Cancer Research, and one of the paper’s senior authors.
The paper’s other senior author is Brian Wolpin, an assistant professor of medical oncology at Dana-Farber. Wolpin, a clinical epidemiologist, assembled the patient sample from several large public-health studies. All patients had their blood drawn when they began participating in the studies and subsequently filled out annual health questionnaires.
Working with researchers at the Broad Institute, the team analyzed blood samples for more than 100 different metabolites — molecules, such as proteins and sugars, produced as the byproducts of metabolic processes.
“What we found was that this really interesting signature fell out as predicting pancreatic cancer diagnosis, which was elevation in these three branched chain amino acids: leucine, isoleucine, and valine,” Vander Heiden says. These are among the 20 amino acids — the building blocks for proteins — normally found in the human body.
Some of the patients in the study were diagnosed with pancreatic cancer just one year after their blood samples were taken, while others were diagnosed two, five, or even 10 years later.
“We found that higher levels of branched chain amino acids were present in people who went on to develop pancreatic cancer compared to those who did not develop the disease,” Wolpin says. “These findings led us to hypothesize that the increase in branched chain amino acids is due to the presence of an early pancreatic tumor.”
Early protein breakdown
Vander Heiden’s lab tested this hypothesis by studying mice that are genetically programmed to develop pancreatic cancer. “Using those mouse models, we found that we could perfectly recapitulate these exact metabolic changes during the earliest stages of cancer,” Vander Heiden says. “What happens is, as people or mice develop pancreatic cancer, at the very earliest stages, it causes the body to enter this altered metabolic state where it starts breaking down protein in distant tissues.”
“This is a finding of fundamental importance in the biology of pancreatic cancer,” says David Tuveson, a professor at the Cancer Center at Cold Spring Harbor Laboratory who was not involved in the work. “It really opens a window of possibility for labs to try to determine the mechanism of this metabolic breakdown.”
The researchers are now investigating why this protein breakdown, which has not been seen in other types of cancer, occurs in the early stages of pancreatic cancer. They suspect that pancreatic tumors may be trying to feed their own appetite for amino acids that they need to build cancerous cells. The researchers are also exploring possible links between this early protein breakdown and the wasting disease known as cachexia, which often occurs in the late stages of pancreatic cancer.
Also to be answered is the question of whether this signature could be used for early detection. The findings need to be validated with more data, and it may be difficult to develop a reliable diagnostic based on this signature alone, Vander Heiden says. However, he believes that studying this metabolic dysfunction further may reveal additional markers, such as misregulated hormones, that could be combined to generate a more accurate test.
The findings may also allow scientists to pursue new treatments that would work by targeting tumor metabolism and cutting off a tumor’s nutrient supply, Vander Heiden says.
MIT’s contribution to this research was funded by the Lustgarten Foundation, the National Institutes of Health, the Burroughs Wellcome Fund, and the Damon Runyon Cancer Research Foundation.