[SYS_OK] MEA_0 ACTIVE -- 862 Hz mean spike rate -- donor 924A consent verified -- 37.0 C -- rack draw 914 W -- 0.85 Gop/W -- MEA_1 ACTIVE -- 743 Hz -- donor 482B consent verified -- 37.1 C -- perfusion nominal --
OrganoidBiofoundry
Drug discovery · PDAC

Pancreatic cancer organoids and the desmoplastic barrier

Pancreatic ductal adenocarcinoma is hard to treat largely because the cancer cells hide inside a dense fibrotic shield. A useful PDAC organoid model has to reproduce that shield, not just the cancer, or it tests the wrong thing.

In a real pancreatic tumor, the cancer cells are often less than a fifth of the volume; the rest is stroma. Any model that screens drugs against the cancer cells alone will flatter compounds that could never reach them in a patient. The work, and the difficulty, is in rebuilding the barrier.

What the desmoplastic barrier is

The pancreatic tumor stroma is a dense, cross-linked extracellular matrix of type I and III collagens, hyaluronic acid, and fibronectin, stiffened by lysyl oxidase cross-linking to an elastic modulus above 10 kilopascals, against under 1 kilopascal for normal pancreas. That stiffness raises interstitial fluid pressure enough to compress local vessels, so standard chemotherapeutics such as gemcitabine struggle to be delivered at all. The barrier is built and maintained by activated pancreatic stellate cells that become cancer-associated fibroblasts.20

Drug penetration across the desmoplastic stroma A gradient from the perfused vessel side through dense myofibroblastic stroma and inflammatory stroma into the tumor core, with drug concentration falling sharply across the stiff matrix. vessel -> core Vessel / perfused edge drug enters here Myofibroblastic stroma (myCAF) dense collagen, stiff Inflammatory stroma (iCAF) cytokines, immune evasion Tumor core often untouched, drives recurrence
Drug concentration falls steeply from the vessel side into the tumor core across the stiff, cross-linked stroma. A high Thiele modulus means the periphery consumes the drug before it reaches the center.

How the model reproduces it

PDAC-on-a-chip systems place a tumor organoid in a tunable hydrogel, often a collagen-hyaluronic acid network whose stiffness is set by the cross-link density, flanked by perfusion channels lined with endothelial cells to mimic vasculature under physiological shear. Co-culturing the organoid with myofibroblastic and inflammatory cancer-associated fibroblasts recreates both the mechanical and the chemical components of the barrier. The point is to control the variables, matrix stiffness, fibroblast density, flow, so that drug penetration can be measured rather than assumed.

Measuring whether the drug gets through

Endpoint viability alone cannot distinguish a drug that failed because the cells resist it from one that simply never arrived. So penetration is measured directly: fluorophore-tagged or label-free imaging traces a compound through the stroma, and the radial concentration is fit to a transport model. The governing balance is convection plus diffusion against cellular uptake, summarized by the Thiele modulus, the ratio of consumption rate to diffusion rate.

The transport balance

Drug flux follows J = -D∇C + vC, where D is the diffusion coefficient in the desmoplastic matrix, ∇C the concentration gradient, and v the convective velocity from interstitial flow. A high Thiele modulus means uptake outruns diffusion: the drug is consumed at the periphery and the core is spared. Mapping how matrix composition changes D gives pharmaceutical partners a kinetic, not just a yes-or-no, readout.

Two-target screening

The model's value is testing combinations: a stromal modulator to open the barrier, paired with a cytotoxic or targeted payload to kill the exposed tumor. Stromal modulators under study include lysyl oxidase inhibitors to block collagen cross-linking, hyaluronidases to degrade hyaluronic acid, and TGF-beta antagonists to deactivate fibroblasts; payloads range from FOLFIRINOX to KRAS-targeted inhibitors. Giving the modulator first or alongside the payload, and watching local impedance drop as the matrix degrades, shows whether opening the stroma actually widens the therapeutic window.21

Frequently asked questions

Why is pancreatic cancer so hard to treat?

Largely because a dense, stiff stroma surrounds the cancer cells, raising fluid pressure, compressing vessels, and physically blocking drugs from reaching the tumor.

What makes a PDAC organoid model useful?

Reproducing the stroma, not just the cancer cells. Co-culture with cancer-associated fibroblasts and a tunable matrix lets a screen test whether a drug can actually penetrate the barrier.

What is the Thiele modulus doing here?

It captures whether drug uptake outruns diffusion. A high value means the periphery consumes the drug before it reaches the tumor core, which is the core failure mode in pancreatic cancer.

References

  1. Ohlund D, et al. Distinct populations of inflammatory fibroblasts and myofibroblastic cancer-associated fibroblasts in pancreatic cancer. Journal of Experimental Medicine. 2017;214(3):579-596. doi:10.1084/jem.20162024. Accessed 2026-06-12.
  2. Tiriac H, et al. Organoid profiling identifies common responders to chemotherapy in pancreatic cancer. Cancer Discovery. 2018;8(9):1112-1129. doi:10.1158/2159-8290.CD-18-0349. Accessed 2026-06-12.