Discovery reveals how drug-resistant tumours can shift into a state that responds better to chemotherapy.
Singapore, 3 March 2026— Scientists at Duke-NUS Medical School have identified a molecular “switch” that determines whether pancreatic cancer cells resist chemotherapy or respond to it—a finding that could help convert some of the most treatment-resistant tumours into forms that are more manageable with existing drugs.
Published in the Journal of Clinical Investigation, the study reveals the molecular mechanism behind this switch, suggesting that combining targeted drugs with standard chemotherapy could improve outcomes for patients with therapy-resistant forms of the disease.
Pancreatic cancer is among the deadliest cancers worldwide. In Singapore, it is the ninth most common cancer but the fourth leading cause of cancer death[1]. Because it is often diagnosed late and responds poorly to treatment, most patients rely on chemotherapy, which typically offers only modest benefits.
Over the past decade, researchers have learned that pancreatic cancers fall into two main molecular subtypes—classical and basal. Cancer cells in the classical subtype are more structured and organised, with patients generally responding better to treatment. In the basal subtype, the cells are more disorderly and aggressive, often resistant to chemotherapy.
Most importantly, cancer cells are not locked into a single subtype. They transition between more treatable and more resistant states, a process known as cancer cell plasticity. The team focused on a gene, called GATA6, which is known to help keep pancreatic cancer cells in the more organised, less aggressive state. When GATA6 levels are high, the cancer tends to grow in a more structured way and responds better to chemotherapy. When GATA6 levels drop, the cancer cells lose this organisation, becoming more aggressive and likely to resist treatment.
Professor David Virshup of Duke-NUS’s Programme in Cancer & Stem Cell Biology, who is lead author of the study, said:
“We have known that pancreatic cancer cells can switch between these two states. What we didn’t understand was the mechanism driving that switch. By identifying the pathway that suppresses GATA6, we now have a clearer picture of how tumours become resistant—and potentially how to reverse that process.”
The researchers identified a chain of signals inside pancreatic cancer cells that controls this process. A gene called KRAS, which is mutated in nearly all pancreatic cancers, tells the cells to keep growing constantly. KRAS sends signals through a helper protein called ERK, which carries these growth instructions deeper into the cell.
When the ERK pathway becomes highly active, it starts protecting another protein which slows down the production of GATA6. With decreased GATA6 levels, pancreatic cancer cells lose their organised structure, becoming more aggressive and much harder to treat with chemotherapy.
Conversely, by using genetic screens, cell-based molecular analyses, and drug treatments in pancreatic cancer cells, the team showed that blocking the KRAS and ERK pathway removes the suppression of GATA6, allowing GATA6 levels to rise. As a result, the cancer cells shift back toward the more organised state and become more sensitive to chemotherapy drugs.
The study also showed that increasing GATA6 levels in general made pancreatic cancer cells more responsive to treatment. When drugs inhibiting the KRAS and ERK pathway were combined with standard chemotherapy, the effects were stronger than either treatment alone, but only when GATA6 was present. This suggests that GATA6 is a key factor in determining when combination therapies are most beneficial.
These findings offer hope for improving treatment strategies for a disease with few effective options. They help explain why patients with higher GATA6 levels tend to respond better to certain chemotherapy regimens, while providing a molecular basis for ongoing clinical trials for novel treatment strategies targeting KRAS and similar pathways.
Professor Lok Sheemei, Duke-NUS' Interim Vice-Dean for Research, said:
“Pancreatic cancer remains one of the toughest cancers to treat. These findings provide a mechanistic explanation for why tumours respond poorly to chemotherapy and offers a rational strategy for combining targeted therapies with existing drugs.”
Beyond pancreatic cancer, the work highlights a broader principle in cancer biology. Many cancers driven by KRAS mutations show similar changes in cell state and treatment response. Understanding how cancer cells switch between different states may help researchers tackle therapy resistance in other cancer types as well.
Professor Patrick Tan, Dean and Provost’s Chair in Cancer and Stem Cell Biology at Duke-NUS, commented:
“This work demonstrates how basic science can uncover actionable insights into treatment resistance. Understanding how cancer cells switch states gives us a more strategic way to design combination treatments.”
Duke-NUS Medical School is a global leader in medical education and a biomedical research powerhouse, combining fundamental discovery with translational expertise to improve health outcomes in Singapore and beyond.