Developing a drug involves establishing an optimal therapeutic index in which the dose administered achieves the desired therapeutic effect, yet has limited safety concerns. Efficacy and safety are each highly dependent on concentrations of the drug that are attained in the organs and tissues of the body over time following administration. The therapeutic effect is determined by the drug concentrations that are attained at the active site of the target, relative to the concentrations required to trigger the desired response in the target. Similarly, adverse events can be caused by off-target effects or accumulation in particular organs or tissues, which again is dependent on localised drug concentrations. Critical side-effects determine the upper limit of safe dosing.
The processes that determine these drug concentrations are known as a drug’s pharmacokinetics (PK). Thus, knowledge of a drug’s PK, in conjunction with relevant target binding data, is critical in the understanding of, and the ability to predict, its therapeutic window. Insight into PK facilitates both the assessment of whether a compound is likely to have an acceptable therapeutic index, and the dose regime(s) that define the therapeutic window.
Having an early insight into the pharmacokinetics can be beneficial in directing synthetic chemistry and prioritising compounds for in vitro screening, with a view to obtaining an optimal therapeutic index. Cyprotex have developed virtual PK prediction software known as chemPKTM which predicts human pharmacokinetics directly from compound structures, requiring no in vitro ADME or physicochemical data and, thus, no compound synthesis. The software utilises a PBPK model optimised on human clinical (in vivo) data for maximum reliability. It is deployed on the easy to use KNIME Analytics Platform.
Details on the model development, optimisation and validation were presented at the 14th European ISSX meeting in Cologne from June 26-29, 2017.