Scientists still do not fully understand how human pancreatic cells first become cancerous, limiting efforts to identify the disease at its earliest stages.
“Early detection of pancreatic cancer is mostly missed because in 80 percent of cases, the tumor becomes symptomatic only after it has progressed to advanced stages,” said Jaskaran Singh Sethi, a radiation oncologist at Rajiv Gandhi Cancer Institute and Research Center, who was not involved in the study.
Together, these challenges have made it difficult to identify early biomarkers or directly study the molecular events that initiate malignant transformation.1
In a new study published in Developmental Cell, researchers developed human pancreatic organoids that reproduced the earliest stages of pancreatic ductal adenocarcinoma (PDAC) development and found that, unlike in many mouse models, activating the cancer-driving gene Kirsten rat sarcoma virus oncogene homologue (KRAS) alone was insufficient to drive tumor formation.2
Instead, the human cells remained resistant unless multiple tumor-suppressive barriers were disrupted, suggesting that early pancreatic tumorigenesis in humans may be more complex than existing animal models have captured. The system also modeled progression toward aggressive forms of PDAC, potentially providing a platform to study early biomarkers and mechanisms of disease progression.
For decades, genetically engineered mouse models have served as the dominant experimental system for studying pancreatic cancer initiation, helping establish oncogenic KRAS as a central driver of PDAC development.3 However, increasing evidence suggested that malignant transformation in human pancreatic cells may require overcoming multiple tumor-suppressive barriers not fully captured in many mouse models.4
The question naturally arose: Which cellular states should scientists study to capture the early footprints of PDAC in humans?
Previous work suggested that pancreatic injury and inflammation can trigger acinar-to-ductal metaplasia (ADM), a regenerative reprogramming process in which mature pancreatic cells adopt a more plastic, progenitor-like state that may become permissive for tumor initiation.5
To model this vulnerable state in a human system, researchers at Memorial Sloan Kettering Cancer Center and Weill Cornell Medicine developed pancreatic organoids from human pluripotent stem cells. The organoids retained embryonic pancreatic progenitor features while also recapitulating molecular characteristics associated with inflamed pancreatic epithelium and ADM.
Using this system, the researchers systematically introduced combinations of mutations commonly found in PDAC to test which genetic alterations were required for transformation.
The researchers found that combining KRAS activation with loss of either the tumor suppressor genes cyclin-dependent kinase inhibitor 2A (CDKN2A) or tumor protein p53 (TP53) still failed to produce tumors after transplantation into mice. However, organoids carrying KRAS activation together with loss of both CDKN2A and TP53 consistently formed pancreatic precursor lesions and moderately differentiated PDAC after transplantation.
In particular, CDKN2A loss appeared necessary for KRAS- and TP53-driven transformation, indicating that human pancreatic cells may possess stronger safeguards against malignant progression than traditional mouse systems have fully captured.
An additional loss of SMAD family member 4 (SMAD4) also drove progression toward a more aggressive, poorly differentiated, basal-like disease state associated with poorer clinical outcomes, suggesting that the organoid system can capture not only early tumor initiation, but also the emergence of clinically aggressive pancreatic cancer.6
The researchers then examined how the organoids changed during malignant progression to better characterize the early molecular landscape of pancreatic tumorigenesis.
They found that this transition involved two parallel shifts: activator protein 1 (AP-1) transcription factors appeared to drive the oncogenic side of this transition, promoting activation of malignant PDAC programs. At the same time, declining expression of ten-eleven translocation methylcytosine dioxygenase 1 (TET1) weakened molecular programs linked to pancreatic lineage identity. The researchers traced both processes to ERK signaling activated by mutant KRAS, suggesting that the same pathway may simultaneously promote cancer-driving programs while dismantling mechanisms that preserve normal cellular identity.
Together, the findings suggest that early pancreatic tumorigenesis may involve not only accumulation of oncogenic mutations, but also progressive erosion of the molecular features that maintain normal pancreatic identity.
By modeling the earliest stages of human pancreatic cancer development, Sethi said, this “important study identifies key molecular and epigenetic events driving [malignant] transformation.”
Such work has the potential to aid early detection, risk prediction, and development of preventive or targeted therapies, he said.
- Lin Z, et al. Early detection of pancreatic cancer: Current advances and future opportunities. Biomedicines. 2025;13(7):1733.
- Zhang X, et al. Human pancreatic progenitor organoids define genetic and epigenetic barriers to early PDAC transformation. Dev Cell. 2026.
- Rozengurt E, Eibl G. Pancreatic cancer: Molecular pathogenesis and emerging therapeutic strategies. Signal Transduct Target Ther. 2026;11:6.
- Stefanoudakis D, et al. Significance of TP53, CDKN2A, SMAD4 and KRAS in pancreatic cancer. Curr Issues Mol Biol. 2024;46(4):2827-2844.
- Marstrand-Daucé L, et al. Acinar-to-ductal metaplasia (ADM): On the road to pancreatic intraepithelial neoplasia (PanIN) and pancreatic cancer. Int J Mol Sci. 2023;24(12):9946.
- Greendyk JD, et al. Association between SMAD4 mutations and GATA6 expression in paired pancreatic ductal adenocarcinoma tumor specimens: Data from two independent molecularly-characterized cohorts. Biomedicines. 2023;11(11):3058.
