Effective treatments for brain tumors remain one of the most urgent and unmet needs in modern oncology. This is due not only to the presence of the neurovascular unit/blood–brain barrier (NVU/BBB) but also to the heterogeneity of barrier alteration in the case of brain tumors, which results in what is referred to as the blood–tumor barrier (BTB). Herein, we discuss this heterogeneity, how it contributes to the failure of novel pharmaceutical treatment strategies, and why a “whole brain” approach to the treatment of brain tumors might be beneficial. We discuss various methods by which these obstacles might be overcome and assess how these strategies are progressing in the clinic. We believe that by approaching brain tumor treatment from this perspective, a new paradigm for drug delivery to brain tumors might be established.

Histone deacetylase expression and activity are often dysregulated in central nervous system (CNS) tumors, providing a rationale for investigating histone deacetylase inhibitors (HDACIs) in selected brain tumor patients. Although many HDACIs have shown potential in in vitro studies, they have had modest efficacy in vivo. This lack of activity could be due to insufficient CNS exposure to the unbound drug. In this study, we investigated the systemic pharmacokinetics and subsequent CNS distribution of 2 potent HDACIs, vorinostat and quisinostat, in the murine model. Both compounds undergo in vitro degradation in mouse plasma, requiring precautions during sample processing. They also have short half-lives in vivo, in both plasma and the CNS, which may lead to diminished efficacy. Transgenic transporter-deficient mouse models show that the CNS delivery of vorinostat was not limited by the 2 major blood-brain barrier efflux transporters, p-glycoprotein and breast cancer resistance protein. Vorinostat had an unbound CNS tissue-to-plasma partition coefficient of 0.06 ± 0.02. Conversely, the exposure of unbound quisinostat in the brain was only 0.02 ± 0.001 of that in the plasma, and the CNS distribution of quisinostat was limited by the activity of p-glycoprotein. To gain further context for these findings, the CNS distributional kinetics for vorinostat and quisinostat were compared with another hydroxamic acid HDACI, panobinostat. A comprehensive understanding of the CNS target exposure to unbound HDACI, along with known potencies from in vitro testing, can inform the prediction of a therapeutic window for HDACIs that have limited CNS exposure to unbound drug and guide targeted dosing strategies. SIGNIFICANCE STATEMENT: This study indicates that quisinostat and vorinostat are susceptible to enzymatic degradation in the plasma, and to a lesser degree, in the target central nervous system (CNS) tissues. Employing techniques that minimize the postsampling degradation in plasma, brain, and spinal cord, accurate CNS distributional kinetic parameters for these potentially useful compounds were determined. A knowledge of CNS exposure, time to peak, and duration can inform dosing strategies in preclinical and clinical trials in selected CNS tumors.