Scientists identify periostin as key to pancreatic cancer's deadly spread

Pancreatic cancer does not spread alone. It rewires the tissue around it, turning healthy cells into accomplices in its own invasion. A team of researchers in Brazil has now identified the mechanism: a protein called periostin, which the tumor uses to reshape the connective tissue surrounding it, carving pathways into nerves that act as highways for cancer cells to reach distant parts of the body. The discovery offers a potential new target for treatment—but it also underscores why this cancer remains so lethal.

Pancreatic adenocarcinoma, which arises in the glandular cells that produce digestive juices, accounts for roughly 90 percent of all pancreatic cancer diagnoses. Globally, about 510,000 people receive this diagnosis each year, and nearly the same number die from it. In Brazil alone, the National Cancer Institute estimates roughly 11,000 new cases annually, paired with 13,000 deaths. The survival numbers are grim: only about one in ten patients lives five years past diagnosis. "It's an aggressive cancer that's difficult to treat," says Pedro Luiz Serrano Uson Junior, an oncologist involved in the research.

One reason for this aggressiveness is a process called perineural invasion—cancer cells that burrow into nerves and travel along them like trains on a track. Because nerves connect different regions of the body, this pathway allows tumors to spread far more efficiently than they could through surrounding tissue alone. Uson describes perineural invasion as "a marker of cancer aggressiveness." More than half of pancreatic cancer patients already show signs of this nerve invasion by the time they are diagnosed, though it is typically discovered only after surgery, when pathologists examine the removed tissue under a microscope. By then, the cancer has already begun its escape.

Researchers at the Center for Research on Inflammatory Diseases in São Paulo, led by Carlos Alberto de Carvalho Fraga and principal investigator Helder Nakaya, set out to understand how this invasion happens at the molecular level. They analyzed 24 pancreatic cancer samples using advanced tools that can track the activity of thousands of genes in individual cells while mapping their precise locations within tumor tissue. What they found was that the stroma—the connective tissue that surrounds and supports the tumor—is not a passive bystander but an active participant in the cancer's spread. Pancreatic and stellate cells within this stroma produce large amounts of periostin, a protein that reshapes the extracellular matrix, the structural scaffold that holds healthy tissue together. "Periostin participates in this remodeling, paving the way for tumor cells to invade," Nakaya explains. Once cancer cells reach a nerve, that nerve becomes a route for further dissemination.

As the tumor environment transforms, it triggers what is known as a desmoplastic reaction—the buildup of dense, fibrous tissue that hardens and inflames the area around the cancer. This protective barrier becomes a problem for treatment. Chemotherapy and immunotherapy drugs cannot easily penetrate this thickened tissue, allowing cancer cells to survive and continue spreading even as medications are administered. "That's why pancreatic cancer is still so difficult to treat," Uson says. The tumor has essentially built a fortress around itself.

The researchers believe periostin represents a promising target for future intervention. Reducing its activity or eliminating the stellate cells that produce it could theoretically limit nerve invasion and slow the cancer's ability to spread before it gains momentum. Clinical trials in other cancers are already testing antibodies designed to block periostin, and those results may inform whether the same approach could work in pancreatic cancer. Uson notes that such a strategy aligns with the emerging field of precision medicine—treating patients based on the molecular and genetic changes in their tumors rather than the tumor type alone. "If we can develop antibodies or drugs that block these stellate cells, we'll have tools to prevent the tumor from acquiring this invasive capacity so early," he says. There is currently no treatment specifically designed to stop perineural invasion, and therapies targeting this process could potentially benefit patients with other aggressive cancers, including those of the intestine and breast.

The next phase of this work involves translating these insights into actual treatments that can act before invasion begins. Nakaya emphasizes that the study also demonstrates the power of analyzing existing public databases with new questions in mind—the researchers were able to uncover patterns that the original data collectors had not considered. The path forward is clear in outline but demanding in execution: precision medicine is advancing, and the future of cancer treatment lies in intervening at the molecular level, before the tumor has time to reprogram its surroundings and escape.