Pancreatic ductal adenocarcinoma (PDAC) is often described as a single, fast-growing mass that spreads early and resists therapy. But a growing body of research is reshaping that picture: many PDACs are better understood as evolving ecosystems made of genetically and transcriptionally distinct clones, distributed across complex ductal and stromal landscapes. New single-cell RNA sequencing and spatial transcriptomics studies published this year extend that view, charting how heterogeneous tumor cell states emerge, compete and cooperate—changes that can influence invasion, metastasis and treatment failure.
At the heart of the problem is clonal heterogeneity: the coexistence of multiple tumor cell populations that differ in driver mutations, gene expression programs and interactions with surrounding tissue. These differences matter because PDAC progression is not governed only by which mutations are present, but also by where clones reside, which microenvironments they occupy, and which phenotypes they adopt as disease advances.
A patchwork tumor, not a uniform mass
Recent transcriptomics work has combined single-cell profiling—which identifies gene expression signatures cell by cell—with spatial methods that preserve each cell’s location in intact tumor sections. Together, these approaches reveal that PDAC tumors can contain multiple malignant cell programs in parallel, including highly proliferative epithelial-like populations, stress-adapted states, and more mesenchymal or invasive phenotypes. Importantly, these programs are not evenly mixed. They appear in “neighborhoods” shaped by ductal anatomy, fibrosis, immune cell infiltration and local nutrient or oxygen availability.
This spatial organization offers a mechanism for how PDAC can progress even when only part of a tumor acquires an especially aggressive state: localized niches can act as incubators for invasion-prone clones. Studies also suggest that transcriptomic states can shift over time, meaning a clone defined by a particular mutation may still diversify into multiple phenotypes depending on microenvironmental pressures.
Evidence for ductal spread and branching evolution
Clonal complexity is also visible when tumors are sampled across multiple sites in the pancreas. In a detailed multi-region sequencing investigation of a pancreatobiliary cancer case, researchers mapped invasive and non-invasive lesions throughout the organ and compared mutation profiles across compartments. The analysis found shared driver alterations—mutations affecting KRAS and TGFBR1—across multiple invasive foci, including lesions involving the distal bile duct. That pattern is consistent with a common ancestral clone seeding several sites rather than each focus forming independently.
At the same time, other non-contiguous, non-invasive lesions carried different KRAS variants, indicating the presence of additional, genetically distinct precancerous clones in the same organ field. Investigators also observed intraductal lesions in the main pancreatic duct with a KRAS G12D mutation paired with an RNF43 alteration, a combination that appeared repeatedly in invasive foci and in additional high-grade dysplasia that later acquired ARID1A changes. The proposed evolutionary narrative was branching: an early “founder” clone established itself in ductal epithelium, then diverged into subclones, one of which progressed toward multifocal invasive disease.
The same case included separate intraductal papillary mucinous neoplasms (IPMNs) with KRAS G12V and GNAS mutations—supporting the idea that a single pancreas may harbor multiple neoplastic lineages simultaneously. Taken together, this and related research backs a model in which tumor cells can disseminate through the ductal system, “skipping” across the gland and generating dispersed lesions that later diverge genetically and phenotypically.
Why heterogeneity accelerates progression
Clonal heterogeneity can speed PDAC progression in at least three ways. First, it creates a reservoir of pre-adapted subclones—cells that already possess traits favoring invasion, immune evasion, or survival under low oxygen and nutrient stress. Second, it enables division of labor: different clones can shape the microenvironment in complementary ways, including remodeling extracellular matrix or altering immune activity, making the tumor as a whole more resilient. Third, it raises the odds that therapy will leave behind a resistant fraction. Even if a treatment targets a dominant clone, minor populations can expand after selective pressure, driving relapse.
Single-cell and spatial data strengthen these concepts by showing how malignant programs align with stromal architecture and immune exclusion. Dense fibrotic regions, for instance, can correlate with immune-poor neighborhoods where certain tumor states thrive, suggesting that physical and immunological barriers can reinforce clonal compartmentalization.
Toward better diagnostics and therapy
This new resource-style mapping of PDAC evolution is not only descriptive. It points to practical challenges for diagnosis: a biopsy from one region may not represent the most dangerous clone elsewhere in the tumor or in the ductal tree. It also supports therapeutic strategies that combine approaches—such as targeting both tumor-intrinsic pathways and microenvironmental features that shelter aggressive states.
Clinically, the findings underscore why “one-size-fits-all” treatment often fails in PDAC. The disease can behave like multiple cancers occupying the same organ, some already poised for dissemination, others feeding progression through stromal and immune modulation. Better risk assessment may require integrating genetics, single-cell phenotypes and spatial context, rather than relying on bulk averages from mixed cell populations.
Why it matters
PDAC remains among the deadliest cancers, and clonal heterogeneity helps explain its rapid progression and frequent treatment resistance. By showing how founder clones branch into distinct lineages and how malignant states are organized in space, single-cell and spatial transcriptomics research provides a roadmap for more representative sampling and more rational combination therapies. Ultimately, understanding “where” aggressive clones arise may be as important as knowing “which” mutations they carry.