Current Topics in Medicinal Chemistry - Volume 11, Issue 4, 2011

Full text available

The use of in vitro metabolism in scaling to predict human clearance of new chemical entities has become a commonplace activity in the research and development of new drugs. The measurement of in vitro lability in human liver microsomes, a rich source of drug metabolizing cytochrome P450 enzymes, has become a high throughput screen in many research organizations which is a testament to its usefulness in drug design. In this chapter, the methods used to scale in vitro intrinsic clearance data to predict in vivo clearance are described. Importantly, the numerous assumptions that are required in order to use in vitro data in this manner are laid out. These include assumptions regarding the scaling process as well as technical aspects of the generation of the in vitro data. Finally, some other drug clearance processes that have been emerging as important are described with regard to ongoing research efforts to develop clearance prediction methods.

This review focuses on a discussion of the controversies in allometric scaling (AS) for predicting human clearance from a mathematical and statistical perspective. First, a history of allometric scaling in comparative biology and its use in pharmacokinetics are reviewed. It is shown that the application of AS in predicting human clearance values based on a limited number of animal species (typically, 3 or 4) contains fundamental statistical errors from when AS was first introduced from comparative biology. Second, the mathematical nature of various allometrically-based methods is revealed and the soundness of these methods is assessed. It is demonstrated that any of these methods, which incorporate a correction factor in a traditional allometric approach (varying-exponent allometry), not only reduces the statistical power of the allometric analysis, but are also incorrect with regard to aspects of biology. Finally, it is concluded that allometry remains a valuable tool for predicting human clearance, and should be applied in the context of a fixed exponent. However, fixed-exponent allometry does not provide satisfactory accuracy in predicting human clearance, since it is not able to capture the biological differences among species. Therefore, it is recommended that the overall effort in predicting human pharmacokinetics should be directed to the collection and generation of reliable data (both in vitro and in vivo) along with a better understanding of the DMPK properties of the chemical entity.

Quantitative human pharmacokinetic (PK) predictions play a critical role in assessing the quality of potential drug candidates and in selecting a human starting dose for clinical evaluation, where the parameters of clearance, volume of distribution, and bioavailability as well as the plasma concentration time profiles are the desired endpoints. While there are numerous reports validating the use of different methods for predictions, it still remains an open question as to what animal species to include when extrapolating the animal PK to human. Given toxicological assessment is generally conducted in two species, a rodent and a non-rodent species, prior to evaluation in human subjects, rat, dog and/or monkey are typically the species ADME scientists employ to evaluate PK. However, the question is, can we achieve an adequate prediction without the use of larger species such as monkey? In the end, the data and tools utilized for human PK predictions will depend on a number of factors such as information from observed human PK for structurally related compounds, the primary mechanism of clearance, and the availability of in silico and in vitro tools applicable to the respective clearance mechanism. Despite these dependencies, for most situations, adequate predictions can be achieved without the use of monkey PK for predicting human.

ADME prediction is an extremely challenging area as many of the properties we try to predict are a result of multiple physiological processes. In this review we consider how in-silico predictions of ADME processes can be used to help bias medicinal chemistry into more ideal areas of property space, minimizing the number of compounds needed to be synthesized to obtain the required biochemical/physico-chemical profile. While such models are not sufficiently accurate to act as a replacement for in-vivo or in-vitro methods, in-silico methods nevertheless can help us to understand the underlying physico-chemical dependencies of the different ADME properties, and thus can give us inspiration on how to optimize them. Many global in-silico ADME models (i.e generated on large, diverse datasets) have been reported in the literature. In this paper we selectively review representatives from each distinct class and discuss their relative utility in drug discovery. For each ADME parameter, we limit our discussion to the most recent, most predictive or most insightful examples in the literature to highlight the current state of the art. In each case we briefly summarize the different types of models available for each parameter (i.e simple rules, physico-chemical and 3D based QSAR predictions), their overall accuracy and the underlying SAR. We also discuss the utility of the models as related to lead generation and optimization phases of discovery research.

Evaluation of the potential of a drug candidate to inhibit or inactivate cytochrome P450 (CYP) enzymes remains an important part of pharmaceutical drug Discovery and Development programs. CYP enzymes are considered to be one of the most important enzyme families involved in the metabolic clearance of the vast majority of prescribed drugs. Clinical drug-drug interactions (DDI) involving inhibition or time-dependent inactivation of these enzymes can result in dangerous side effects resulting from reduced clearance/increased exposure of the drug being affected (the ‘victim’ drug). In this regard, pharmaceutical companies have become quite vigilant in mitigating CYP inhibition/inactivation liabilities of drug candidates early in Discovery including continued risk assessment throughout Development. In this review, common strategies and decision making processes for the assessment of DDI risk in the different stages of pharmaceutical development are discussed. In addition, in vitro study designs, analysis, and interpretation of CYP inhibition and inactivation data are described in stage appropriate context. The in vitro tools and knowledge available now enable the Discovery Chemist to place the potential CYP DDI liability of a drug candidate into perspective and to aid in the optimization of chemical drug design to further mitigate this risk.

A primary objective of pharmacokinetic-pharmacodynamic (PKPD) reasoning is to identify key in vivo drug and system properties, enabling prediction of the magnitude and time course of drug responses under physiological and pathological conditions in animals and man. Since the pharmacological response generated by a drug is highly dependent on the actual system used to study its action, knowledge about its potency and efficacy at a given concentration or dose is insufficient to obtain a proper understanding of its pharmacodynamic profile. Hence, the output of PKPD activities extends beyond the provision of quantitative measures (models) of results, to the design of future protocols. Furthermore, because PKPD integrates DMPK (e.g. clearance) and pharmacology (e.g. potency),it provides an anchor point for compound selection, and, as such, should be viewed as an important weapon in medicinal chemistry. Here we outline key PK concepts relevant to PD, and then consider real-life experiments to illustrate the importance to the medicinal chemist of data obtained by PKPD. Useful assumptions and potential pitfalls are described, providing a holistic view of the plethora of determinants behind in vitro-in vivo correlations. By condensing complexity to simplicity, there are not only consequences for experimental design, and for the ranking and design of compounds, but it is also possible to make important predictions such as the impact of changes in drug potency and kinetics. In short, by using quantitative methods to tease apart pharmacodynamic complexities such as temporal differences and changes in plasma protein binding, it is possible to target the changes necessary for improving a compound's profile.

Metabolic activation of new chemical entities to reactive intermediates is routinely monitored in drug discovery and development. Reactive intermediates may bind to cellular macromolecules such as proteins, DNA and may eventually lead to cell death via necrosis, apoptosis or oxidative stress. The evidence that the ultimate outcome of metabolic activation is an adverse drug reaction manifested as in vivo toxicity, is at best circumstantial. However, understanding the process of bioactivation of structural alerts by trapping the reactive intermediates is critical to guide medicinal chemistry efforts in quest for safer and potent molecules. This commentary provides a brief introduction to adverse drug reactions and mechanisms of reactive intermediate formation for various functional groups, followed by a review of chemical design approaches, examples of such strategies, possible isosteric replacements for structural alerts and rationalization of laboratory approaches to determine reactive intermediates, as a guide to today's medicinal chemist.

It is a commonly accepted assumption that only unbound drug molecules are available to interact with their targets. In order to achieve high unbound plasma drug concentration, it seems obvious to design compounds with low plasma protein binding. Similarly to achieve high unbound tissue concentration, we apparently need compounds with low tissue binding. Our theoretical analysis and experimental data demonstrate that unbound plasma concentration is not determined by plasma protein binding but by hepatic intrinsic clearance after oral dose, and unbound tissue concentration is not determined by tissue binding but determined by unbound plasma concentration and transport properties at the blood-tissue barrier. Reduction of plasma and tissue protein binding for a compound will increase the unbound concentration in vitro but may not increase its unbound plasma or tissue concentration in vivo after oral administration. We conclude that plasma protein and tissue binding are essential parameters to understand pharmacokinetics and pharmacodynamics but they should not be optimized independently in drug discovery. Instead we should focus on reducing clearance and efflux at the blood-tissue barrier to increase in vivo plasma and tissue unbound concentration.

The gradual alignment with all of drug metabolism with all aspects of drug discovery and development has led to a complete realignment of the way the work is conducted. From a background of conducting bespoke in vivo studies much of the work is now in a high throughput screening mode. Large technological advances have been made, but the nature of drug metabolism processes, being multi-system and promiscuous means that much of the help provided to the medicinal chemistry is reactive rather than based on fundamental disposition structure-activity relationships. Lessons learned around the chemical and physicochemical properties more often associated with succesfull discovery and development projects are only moderately helpful when the high value pharmacological targets of today only yield potent ligands outside of the boundaries these properties describe. Pivotal to the impact of these properties is the intrinsic permeability of a molecule, something not as widely recognised as perhaps it should be. Metabolic lability is still a problem and the tactics employed are unchanged in 20 years: attempt to lower lipophilicity, if it is too high overall or introduce blocking groups, particularly halogens, after identifying the sites of metabolism Perhaps the greatest success drug metabolism science has had over the last fifteen years is it's pivotal role in characterising drug-drug interactions and providing screening systems and computational models to investigate them. It still has many undeveloped areas of the science such as the role of metabolites in drug activity and why compounds vary in their extent of biliary excretion.