Combining imaging strategies with physiologically-based pharmacokinetic modeling to improve precision dosing & drug delivery in brain tumors.

Small molecule targeted therapies frequently fail to control brain metastases and primary brain tumors even when systemic exposure reaches toxicity-limiting levels. Systemic exposure-response relationships are often absent, which is well documented for tyrosine kinase inhibitors. Target-site information is therefore needed, since the concentration reached in the tumor is determined not by plasma exposure, but by passive and active transport across the blood-brain barrier and by intratumoral heterogeneity. Physiologically-based pharmacokinetic (PBPK) modeling provides the mechanistic framework to describe this. Models built solely on in vitro-to-in vivo extrapolation of permeability and transport have repeatedly required correction factors derived from rodent data, and their evaluation in humans still depends on invasive sampling or microdialysis. Microdialysis performs poorly for lipophilic, highly protein-bound drugs. We argue that quantitative imaging can supply the missing in vivo input. In a hybrid strategy, in vitro-derived transport parameters serve as the starting point, while blood-brain barrier influx and efflux are recalibrated on positron emission tomography (PET) microdose measurements in patients. This could lead to improved prediction of brain concentrations at therapeutic doses and allows simulated target-site exposure to be weighed against in vitro potency for compound and dose selection. Beyond whole-brain averages, magnetic resonance imaging (MRI) can provide permeability and perfusion parameters longitudinally and in larger cohorts, whereas mass spectrometry imaging (MSI) techniques can map intratumoral heterogeneity in drug distribution in the tumor microenvironment and identify potential sanctuary sites. We map each modality onto the specific PBPK parameters it can inform, and argue that imaging-informed, spatially resolved PBPK is a realistic route toward model-informed precision dosing in neuro-oncology.