How does pancreatic cancer shift toward a more aggressive state? A single-cell study points to MYC

Pancreatic cancer is difficult to treat, and part of the problem may lie within a single tumor: different groups of cancer cells evolve independently, gradually developing features that favor resistance to treatment or spread beyond the tumor. A large-scale single-cell study examined these differences individually and identified a potential point of intervention—blocking MYC activity could shift cancer cells in experiments toward a less aggressive state.

Led by a team at NYU Langone Health and published on October 13 in *Nature Genetics*, the study focused on pancreatic ductal adenocarcinoma. Researchers analyzed 443,451 cells from 62 tumor samples collected from 56 patients; six of those patients provided samples before and after chemotherapy. Reading gene expression in individual cells helps identify differences that analyses of whole tissue samples can obscure, but these nearly 450,000 cells were not all cancer cells.

The study identified multiple coexisting “clones” within tumors—groups of cancer cells with similar genetic characteristics. Gene expression differed across these groups, highlighting the limitations of the “classical” and “basal” classifications commonly used for pancreatic cancer: a continuum of states exists between the less aggressive classical type and the more aggressive basal type, rather than two fixed, mutually disconnected categories.

MYC is a key clue in one of these transition pathways. An earlier preprint record of the study reported that clonal changes from classical to basal states were associated with increased MYC gene copy number and weaker immune response programs. According to a study description released by NYU and published by Medical Xpress, the team further blocked MYC activity in experimental basal-type tumor cells, shifting them toward a classical state and increasing opportunities for contact with immune cells. This suggests that they might become more susceptible to immune attack.

Where cells are located also shapes researchers’ understanding of their behavior. The early preprint indexed in PubMed incorporated 53,236 spatial transcriptomic measurement points and found that more widely dispersed clones showed stronger signals of epithelial–mesenchymal transition and immune suppression. These features were associated with metastatic potential and colonization of lymph nodes, linking the cells’ molecular states to their distribution within tissue. However, correlation alone cannot establish that every clone will cause metastasis.

This atlas still has temporal and evidentiary limitations. Only six patients had paired samples from before and after chemotherapy, and most of the cell data reflect the state at the time of sampling, leaving evolutionary relationships to be inferred through analysis. The PubMed page records a 2025 preprint and explicitly states that it had not yet undergone peer review at that time; it should not be conflated with the version now published in the journal.

The changes induced by MYC blockade also remain experimental evidence. A shift in cancer cells toward a less aggressive expression state does not mean that patients’ tumors have shrunk, nor has it been shown to extend survival. The study raises more specific treatment questions: can cancer cells be prevented from shifting into states that are difficult to treat, and can opportunities for immune attack be improved? Answering these questions will require further verification of whether the effects are lasting, which tumors they apply to, and the safety of targeting MYC.