A Single-Patient NF1 Multi-Organoid System Combines Drug Disposition with Tumor Response
An NF1-mutant breast-cancer model links patient-derived intestinal, liver and kidney tissues with tumor and normal spheroids. The system compares drug handling with tumor effects and tests NF1 exon skipping, but lacks human pharmacokinetic or treatment-outcome validation.
Building a Patient Background Across Several Tissues
A drug’s effect on an isolated tumor can differ from its effect after absorption and metabolism. Researchers at the Korea Research Institute of Bioscience and Biotechnology and collaborating academic centers addressed this problem with a multi-organoid system for one patient with NF1-mutant breast cancer. The study, published in Signal Transduction and Targeted Therapy, combined patient-derived tissue models with drug-concentration and response measurements.
Fourteen breast biospecimens entered the initial culture effort, and five maintained tumor cultures beyond passage five. The personalized experiments centered on one selected patient. Induced pluripotent stem cells made from that patient’s normal fibroblasts carried an NF1 exon-2 frameshift. The tumor also carried PIK3CA and HER2 variants, so its behavior cannot be attributed to NF1 alone. A single healthy fibroblast-derived iPSC background supplied the main comparison.
Linked Compartments Added Absorption and Metabolism
The system connected intestinal organoid-derived epithelia, liver organoids, kidney organoid-derived renal epithelia and a chamber for tumor or matched normal multicellular spheroids. The intestinal and renal components were epithelial layers derived from organoids. The tumor spheroids came from patient tissue, separately from the iPSC-derived organ models.
Patient-derived liver tissues showed reduced albumin and metabolic or transporter functions, together with mesenchymal features. Intestinal barrier resistance differed from the healthy reference, although both backgrounds remained below the Caco-2 reference. The linked system maintained structure over four days. These measurements established donor-dependent differences within immature tissue models.
For pharmacokinetic experiments, 10 μM compounds entered through the intestinal apical side or the metabolism compartment. Liquid chromatography–tandem mass spectrometry followed concentrations for up to 72 hours. Comparison of the healthy system with mouse data provided an external reference; patient plasma pharmacokinetics were not measured.
Drug Disposition Changed the Ranking Suggested by Direct Exposure
Direct tumor tests favored gedatolisib, with an IC50 of 6.1 μM, over bimiralisib at 15.7 μM and paxalisib at 19.4 μM. Absorption and metabolism measurements changed the practical comparison. Everolimus showed negligible absorption in this system, while alpelisib also showed poor absorption. The investigators then compared the disposition of three PI3K/mTOR-pathway inhibitors, including paxalisib.
Paxalisib’s performance in the linked system supported its selection for further experiments despite its weaker direct potency. Its estimated area under the concentration–time curve was 1,553.97 nmol/L·h in the patient system and 1,140.9 nmol/L·h in the healthy reference; the patient-system half-life was 56.93 hours.
NF1 Exon Skipping Altered Tumor Sensitivity
The team used a lentiviral U7 small nuclear RNA system to express antisense sequences that promoted skipping of NF1 exon 2. Selected sequences restored NF1 protein expression and reduced PI3K–AKT signaling in the experimental models. This intervention changes RNA processing; it does not correct the genomic mutation. In a separate response experiment, exon skipping shifted paxalisib IC50 from approximately 16 to 1.2 μM.
After 10 μM paxalisib for 72 hours, intestinal, liver and kidney tissue viability did not fall significantly. Intestinal and renal electrical barrier resistance nevertheless decreased, and stress-associated markers increased. Viability alone therefore gave an incomplete account of the response. Intracellular drug concentrations did not differ significantly between antisense and control conditions, while tumor killing was stronger after NF1 exon skipping, supporting a change in cellular sensitivity.
Preclinical Comparisons Leave Human Validation Open
Exon skipping plus paxalisib also suppressed growth in a mouse xenograft experiment, without significant changes in body weight or the reported major-organ histology. Human efficacy and tolerability were not assessed in this experiment.
The system’s principal contribution is to connect tissue-specific drug handling with tumor response in one genetic background. Interpretation remains constrained by one patient, a single healthy reference, immature organ models and absent vascular and immune compartments. Human pharmacokinetic measurements and prospective treatment outcomes would be needed to test whether the platform’s rankings or exposure estimates transfer to patients.