Identification of Novel, Selective Ataxia-Telangiectasia Mutated Kinase Inhibitors with the Ability to Penetrate the Blood−Brain Barrier: The Discovery of AZD1390
Achieving CNS penetration, when desired, is a hugely important and sometimes very difficult hurdle to overcome. To do this, the Blood-Brain Barrier (BBB): with its unique lining of endothelial cells and its cocktail of efflux pumps / transporters, must be traversed efficiently. Only a small fraction of drugs actually do this. Guidelines to aid the development of CNS-penetrant compounds have been described, such as: minimising molecular weight, n rotatable bonds, n HBD, tPSA. Passive permeability must be maximised, with an efflux ratio of <2 to stand a good chance of crossing into the CNS.
In this paper, the discovery of AZD1390 is described, wherein CNS penetration is essential to serve the target indication: Glioblastoma Multiforme, which is the most common and lethal form of brain tumour. The molecular target is Ataxia-Telangiectasia Mutated (ATM), part of the PI3K related kinase family, whose function includes the early detection of- and protection against DNA double strand breaks (DSBs). Once activated by DSBs, ATM phosphorylates downstream targets such as γH2AX, p53-MDM2 and Chk2, determining cell survival or cell death. Activation of these pathways within glioma cells results in radio-resistance, so part of the therapeutic strategy with AZD1390 is to cause these cancerous cells to remain susceptible to radiotherapy.
AZD1390 is currently at the stage of expanded Phase I clinical trials for treating glioma patients in combination with radiotherapy.
Figure 1: Optimisation path to AZD1390
The authors began with their previous ATM inhibitor, AZD0156, which was potent and selective with great PK, but an MDCKII-MDR1-BCRP efflux ratio of 23.4 meant that it was unlikely to achieve the CNS exposure required for treating intracranial malignancies. A truncated analogue, 5, which was less potent and selective than AZD0156 but had lower efflux and a respectable unbound brain-to-plasma ratio (Kp,uu) of 0.67, was taken as the start point.
Careful navigation of property space, modifying nHBD, rigidity (n rotatable bonds), basicity etc., and known pharmacophores from previous work led to the candidate compound: AZD1390 with the desired profile for the best chance of getting sufficient CNS exposure. One interesting challenge encountered during optimisation, was the observation of rodent-specific efflux, indicated by low Kp,uu (rat and mouse, brain slice method) values despite a MDCKII-MDR1-BCRP efflux ratio of <2. The rodent specific efflux effects were confirmed by seeing a massive increase in Kp,uu after co-dosing with elacridar (efflux pump inhibitor), and by measuring a rat-specific efflux ratio in an LLC-PK1-rMdr1a assay of 3.2.
PET imaging and quantification studies were then undertaken in cynomolgus monkeys, using [11C] radiolabelled compounds. This data produced incredible cross-sectional images, showing [11C]-AZD1390 entering the brain while [11C]-AZD0156, which is highly effluxed, did not. This data also allowed the authors to determine the Kp,uu of AZD1390, in the cynomolgus monkey by obtaining the partition coefficient.
Identification of Novel, Selective Ataxia-Telangiectasia Mutated Kinase Inhibitors with the Ability to Penetrate the Blood−Brain Barrier: The Discovery of AZD139