Hyaluronan-Supported Peritoneal Tumor Explants Preserve Tissue Context and Ascites-Dependent Drug Responses

Preserving Resected Tissue for Drug Evaluation

Researchers at the National University of Singapore, National Cancer Centre Singapore and collaborating institutions developed a hydrogel-supported culture of peritoneal-metastasis tumor explants. Fresh tissue was collected from 21 patients over two years, encompassing several primary tumor types. Individual preservation, profiling and drug experiments used separate subsets of those samples.

The model retained pieces of resected tissue rather than reconstructing a tumor from isolated cell populations. Precision-cut slices were 300 μm thick and were compared on polytetrafluoroethylene (PTFE) inserts, in Matrigel or embedded in hyaluronan-based hydrogels. The material included crosslinks that can be cleaved by matrix metalloproteinases and optional integrin-binding motifs.

An initial 1.5% hyaluronan formulation improved slice-area retention, viability and histology compared with the conventional culture conditions. Further tuning identified 2% hyaluronan, with a storage modulus around 350 Pa, as the best-performing tested formulation. Stiffer 3% and 4% gels preserved area well but supported poorer viability. Tissue size alone therefore did not establish tissue health.

In the initial 1.5% formulation, approximately 60% of cells remained viable at day 12. In the degradable material comparison, hydrogels lost up to 41% of their mass over 12 days while still supporting the tissue. Matrix persistence and tissue preservation therefore were related design questions rather than the same endpoint.

Embedding and Matrix Chemistry Have Different Effects

Stiffness-matched comparisons separated material properties. Degradable hydrogels preserved viability better than non-degradable counterparts. Hyaluronan also outperformed a polyethylene-glycol comparison despite similar area retention. By contrast, replacing the RGD adhesion motif with the scrambled RDG sequence did not significantly change the tested preservation outcomes.

Tissue contraction provided a mechanistic clue. Slices on flat supports shrank in multiple dimensions and showed collagen remodeling, whereas three-dimensional embedding limited further contraction. Short early treatment with inhibitors affecting myosin-II-mediated contractility also reduced contraction and improved day-7 viability. The authors link these observations to a wound-like myofibroblast response, while acknowledging possible off-target effects of the inhibitors.

Day-8 single-cell profiling in three patient samples found retention of major tumor, fibroblast, immune and endothelial populations. Basic cell-type proportions did not differ clearly between hyaluronan and conventional matrices, despite their viability differences. Better preservation was more apparent in selected T-cell and macrophage gene programs and in inferred intercellular communication. These transcript-based results do not directly measure preserved immune killing.

Ascites Changes Drug Response More Than Tissue Survival

The drug assay used a baseline metabolic measurement followed by 48 hours of exposure. In the ascites experiments, explants were pretreated with medium containing 5% patient ascites for 24 hours. Cisplatin, doxorubicin and erlotinib produced patient- and dose-dependent responses; one explant sample remained relatively insensitive to cisplatin even at 80 μM.

Ascites did not significantly improve slice-area retention, viability, histology or cell-type composition compared with serum-containing conditions. It nevertheless increased tolerance to the tested drugs. Single-cell comparisons in two patient samples showed accompanying inflammatory, angiogenesis-associated and immunosuppressive expression changes, including altered T-cell, macrophage and fibroblast programs.

This separation matters for drug evaluation: two cultures can appear similarly preserved while yielding different treatment responses because their fluid environment differs. The study measures that effect in the explant setting rather than assuming that an acceptable viability window guarantees equivalent pharmacology.

A Fluid-Dependent Target with Important Exceptions

The team then tested TM5441, a plasminogen activator inhibitor-1 inhibitor, using six ascites samples with different PAI-1 levels. Responses varied with both tumor source and drug concentration and were associated with ascites PAI-1. However, two ascites samples did not sensitize the tested tumors to inhibition, preserving a more selective response pattern than earlier simpler cultures had suggested.

The authors compare this pattern with their previously published mouse work; those animal results are separate evidence. The present study supports an explant framework for testing how intact tissue and patient fluid jointly influence drug response. Preservation for up to 12 days remains shorter than many clinical treatment cycles, and systemic delivery, vascular exposure and prospective prediction of patient outcomes were not established.