Pre-clinical Modeling of Navtemadlin Pharmacokinetics, Pharmacodynamics, and Efficacy in IDH-wildtype Glioblastoma
Clinical Cancer Research. 2025;31(17):3771-3786. DOI 10.1158/1078-0432.CCR-25-0244.
How to cite
AMA
Vaubel RA, Zhang W, Oh JH, Mladek AC, Pasa TI, Gantchev JK, et al. Pre-clinical Modeling of Navtemadlin Pharmacokinetics, Pharmacodynamics, and Efficacy in IDH-wildtype Glioblastoma. Clin Cancer Res. 2025;31(17):3771-3786. doi:10.1158/1078-0432.CCR-25-0244APA
Vaubel, R. A., Zhang, W., Oh, J. H., Mladek, A. C., Pasa, T. I., Gantchev, J. K., et al. (2025). Pre-clinical Modeling of Navtemadlin Pharmacokinetics, Pharmacodynamics, and Efficacy in IDH-wildtype Glioblastoma. Clinical Cancer Research, 31(17), 3771-3786. https://doi.org/10.1158/1078-0432.CCR-25-0244BibTeX
@article{vaubel2025preclinical,
title = {Pre-clinical Modeling of Navtemadlin Pharmacokinetics, Pharmacodynamics, and Efficacy in IDH-wildtype Glioblastoma},
author = {Vaubel, R. A. and Zhang, W. and Oh, J. H. and Mladek, A. C. and Pasa, T. I. and Gantchev, J. K. and others},
journal = {Clinical Cancer Research},
volume = {31},
number = {17},
pages = {3771--3786},
year = {2025},
doi = {10.1158/1078-0432.CCR-25-0244}
}Glioblastoma is hard to treat in part because the blood–brain barrier keeps many otherwise-promising drugs out of the tumor. Navtemadlin blocks the MDM2 protein to reactivate the p53 tumor-suppressor pathway, and an early-phase trial had already measured the drug reaching patient tumors — but how much drug is actually enough was still unclear.
The team tested navtemadlin across a panel of patient-derived glioblastoma models, tying measured drug levels to response. The strongest, most durable responses came in tumors carrying extra copies of MDM2, and benefit in the brain was throttled by drug efflux at the blood–brain barrier.
Building a model of the exposure needed for benefit, they estimated that only a minority of early-phase patients had reached it, giving a framework for reading those clinical results.
Key findings
- Navtemadlin efficacy tracked with MDM2 amplification across a 15-line glioblastoma xenograft panel. In the single-animal screen, all four MDM2-amplified models showed durable responses that extended time to endpoint 20.1- to 27.8-fold over vehicle, while nine MDM2-non-amplified models reached only 1.9- to 4.8-fold.
- Benefit in the brain was limited by drug efflux at the blood–brain barrier. Navtemadlin's brain-to-plasma ratio was 0.009 in normal mice, and in efflux-transporter-knockout mice a 25 mg/kg dose extended median survival of orthotopic tumors to 132 days versus 62 days for vehicle.
- Spatially registered imaging linked intratumoral drug levels to p53-pathway activation. Cyclic immunofluorescence maps of p53, p21, MDM2, and cleaved caspase-3 were co-registered with mass-spectrometry drug-distribution images of the same brain-tumor sections, where a median tumor navtemadlin concentration of 630 nmol/L coincided with upregulated p53-pathway markers.
- A translational PK/efficacy model set a minimum effective tumor exposure that most early-phase patients did not reach. The modeled target exposure was an AUC of 2,800 hour*ng/mL, and only four of 16 tumor samples at the 240 mg dose level reached it.
CyteFinder II in the methods
“t-CyCIF images were acquired on a CyteFinder II slide scanning fluorescence microscope (RareCyte Inc.) with CyteFinder software (v3.11.024). Image processing and data quantification were performed as previously described (27). Stitching and registration of tiles and cycles were done in MCMICRO (28) using the ASHLAR (v1.10.2; ref. 29) module. Additional details can be found at www.cycif.org. Cell segmentation and marker quantitation were performed with QuPath (25).”
— Vaubel et al., Clinical Cancer Research (2025), Methods, “t-CyCIF”
Why it matters for CyteFinder II users
If you are weighing the CyteFinder II for a tissue study, look at the one job its imaging was asked to do here. The paper's pharmacokinetic and efficacy findings — the brain-to-plasma ratio, the survival curves, the exposure threshold — came from mass spectrometry, dosing studies, and modeling, not from the microscope. What the CyteFinder II carried was the spatial question: after navtemadlin reached the tumor, where in the tissue did the p53 pathway actually switch on? The instrument acquired multi-cycle t-CyCIF images of p53, p21, MDM2, Ki67, and cleaved caspase-3 across whole fresh-frozen brain-tumor sections, one staining round at a time. Those images were then registered to the matching drug-distribution maps of the same tissue, so drug level and pathway response could be read in one shared set of coordinates. That is what the CyteFinder II adds to a pharmacology study: it turns a bulk tissue readout into a spatial one, and it leaves the stitching, segmentation, and quantitation steps open for you to choose.
