Thwip. That, I imagine, is the sound of a tiny pancreatic duct cell slipping into its precancer jacket while the surrounding tissue is still looking for its reading glasses.
Pancreatic cancer has always had a grim habit of arriving late to its own diagnosis. By the time it announces itself, it has often already redecorated the place, recruited unhelpful neighbors, and made treatment feel like trying to evict a tenant who has changed the locks. The new study by Elhossiny and colleagues in Cancer Discovery asks a beautifully old-fashioned question with very modern tools: what happens before the cancer is fully cancer? More precisely, what happens to PanINs, those microscopic pancreatic intraepithelial neoplasias that can precede pancreatic ductal adenocarcinoma, and to the tissue around them? [1]
The Usual Suspects, Now With Seating Charts
For decades, cancer biologists have stared at tumor cells the way village constables in old novels stared at footprints. The cancer cell was the culprit. Find its mutations. Track its escape routes. Interrogate KRAS under a bare light bulb.
That was not wrong. KRAS mutations are practically part of the pancreatic cancer family crest. But tumors do not grow in empty space. They grow in neighborhoods filled with fibroblasts, immune cells, blood vessels, matrix, chemical signals, and assorted biological busybodies collectively known as the tumor microenvironment. If the cancer cell is the bad actor, the stroma is the stage crew, zoning board, security detail, and sometimes the getaway driver.
This study used donor pancreata and spatial technologies to compare normal pancreas, PanIN lesions, and pancreatic cancer. Spatial transcriptomics is, in plain English, a way to ask not only “which genes are active?” but also “where exactly is that happening?” It is molecular biology with a seating chart, which is a mercy after years of grinding tissues into soup and then acting surprised that nobody remembered where they had been sitting.
The Plot Twist Is Timing
Here is the central finding: the epithelial cells in PanINs already look like they are moving along a continuum toward cancer, but the surrounding stroma does not keep pace.
That is the “asynchronous evolution” in the title. The epithelial compartment and the stromal compartment are not aging into cancer together like a synchronized swimming team. The epithelial cells are inching toward malignancy, while the microenvironment in PanINs remains strikingly different from the one seen in full pancreatic cancer.
In cancer, the researchers saw major geographic rearrangements of myeloid cells and lymphocytes, the immune system’s assorted guards, scouts, and sometimes tragically confused hall monitors. They also found a cancer-specific fibroblast population marked by high smooth muscle actin, LRRC15, and LEF1, a WNT signaling component. Fibroblasts have had a long career in pancreatic cancer biology, often cast as the dense, fibrotic wall that makes pancreatic tumors so difficult to treat. LRRC15-positive fibroblasts, in particular, have drawn attention as potentially immunosuppressive stromal players in several cancers. [2]
The important point is not merely that cancer has a strange neighborhood. We knew that. The important point is that many PanINs may fail to become dangerous because their neighborhood never gets properly remodeled.
Why That Matters
This helps explain a puzzle that has bothered the field for years: PanINs are common, pancreatic cancer is not. Recent 3D mapping work from Johns Hopkins found that normal adult pancreata can contain many PanINs, often with oncogenic KRAS mutations, yet only a fraction ever become invasive cancer. [3] If mutations alone were destiny, our pancreata would be far more dramatic organs, and frankly they already have enough responsibilities.
Elhossiny and colleagues suggest that progression may require more than a rebellious epithelial clone. The surrounding tissue may need to be persuaded, bullied, bribed, or otherwise reprogrammed into supporting invasion. That makes the stroma less like wallpaper and more like a second clock. Cancer may need both clocks to strike midnight.
This also fits with recent spatial studies showing that pancreatic tumors are not uniform lumps but organized ecosystems, with treatment and local cell-cell interactions reshaping immune and stromal communities. Hwang and colleagues mapped multicellular dynamics associated with neoadjuvant treatment, while Shiau and colleagues showed therapy-associated remodeling of the pancreatic tumor microenvironment at spatial resolution. [4,5] The field has moved from “what cells are present?” to “who is standing next to whom, whispering what?”
The Long View From the Lab Bench
Those of us who remember when interferon was going to cure everything have learned to respect the distance between a fine biological insight and a useful clinical tool. This study does not mean we can yet scan every pancreas, point at a PanIN, and declare its future intentions like a fortune teller with a pipette.
But if these findings hold up, they may sharpen early detection and interception strategies. Instead of looking only for suspicious epithelial changes, researchers might search for stromal reprogramming: immune cell rearrangements, LRRC15-positive fibroblast neighborhoods, or WNT-associated fibroblast states that mark lesions becoming more dangerous. That could help distinguish the PanINs that are merely loitering from the ones assembling a full criminal enterprise.
It may also influence therapy. Pancreatic cancer has been painfully resistant to immunotherapy, partly because its microenvironment often keeps T cells at a polite but useless distance. Recent reviews emphasize that targeting the pancreatic cancer microenvironment remains difficult, but increasingly necessary. [6] If stromal conversion is a gatekeeper event, then blocking that conversion could become a form of cancer prevention rather than late rescue. Prevention, I should add, is much nicer than rescue. Rescue usually involves more paperwork.
A Small Lesion With a Big Lesson
The charm of this paper is that it restores sequence to a disease we usually meet too late. It says: do not look only at the aspiring cancer cell. Look at the room. Look at the neighbors. Look at whether the immune guards have been moved to the wrong hallway and whether fibroblasts have started wearing the suspicious little LRRC15 name tag.
Pancreatic cancer remains one of oncology’s hardest problems. But studies like this give us a better map of the prehistory. And in cancer, as in archaeology, prehistory is where the important clues often sit quietly, waiting for someone patient enough to brush away the dust.
References
-
Elhossiny AM, Kadiyala P, Okoye JO, et al. Asynchronous evolution of epithelium and stroma differentiates precursor lesions from pancreatic cancer. Cancer Discovery. 2026. DOI: 10.1158/2159-8290.CD-25-2001. PMID: 42165710.
-
Dominguez CX, Müller S, Keerthivasan S, et al. LRRC15+ myofibroblasts dictate the stromal setpoint to suppress tumour immunity. Nature. 2022. PMCID: PMC9630141.
-
Braxton AM, Kiemen AL, Grahn MP, et al. 3D genomic mapping reveals multifocality of human pancreatic precancers. Nature. 2024;629:679-687. DOI: 10.1038/s41586-024-07359-3.
-
Hwang WL, Jagadeesh KA, Guo JA, et al. Single-nucleus and spatial transcriptome profiling of pancreatic cancer identifies multicellular dynamics associated with neoadjuvant treatment. Nature Genetics. 2022;54:1178-1191. DOI: 10.1038/s41588-022-01134-8.
-
Shiau C, Cao J, Gong D, et al. Spatially resolved analysis of pancreatic cancer identifies therapy-associated remodeling of the tumor microenvironment. Nature Genetics. 2024;56:2466-2478. DOI: 10.1038/s41588-024-01890-9. PMCID: PMC11816915.
-
Kung HC, Zheng KW, Zimmerman JW, et al. The tumour microenvironment in pancreatic cancer - new clinical challenges, but more opportunities. Nature Reviews Clinical Oncology. 2025;22:969-995. DOI: 10.1038/s41571-025-01077-z.
Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.