CPSA Digest 2001

Proceedings -Wednesday, October 10, 2001

WOE1

Desirable ADME Properties:
What Are They and How to Screen for Them?

Background

The market for new medicines today is very competitive and companies can gain market share for a superior product simply because one drug has more desirable ADME properties than another (e.g., does not inhibit cyp450 isoenzymes, is metabolized through a different CYP route, or has a longer half-life or a more complete absorption). Screening for these superior ADME properties in potential drug candidates is considered a bottleneck in drug discovery and development for the following reasons:

• There are more compounds to screen
• There is more structural diversity within a drug class
• There are a greater number of therapeutic targets to assess
• There is a furious race to get a drug to market faster than ever before

Discovering poor PK properties as early in the development process as possible would be valuable in reducing failures of new chemical entities. Thus, the trend is to screen for PK properties early in the drug discovery process and to make this approach predictive and successful, and to do so in a high throughput manner.

Premise

Drug Metabolism departments have begun to meet the challenge for early identification of leads/failures by implementing ADME screening tools and strategies early in the drug discovery process. In order to guide useful studies, it is important to identify what the "desirable" ADME properties are.

Desirable ADME properties include the following:

• Good aqueous solubility
• Linear PK for the intended route of administration
• Balanced clearance in terms of renal excretion, biliary secretion, and minimal metabolism with minimal species differences
• No chemically reactive metabolites
• Oxidative metabolism catalyzed by several P450s
• Oxidative metabolism not dependent on polymorphically expressed P450s
• Minimal P450/Pgp inhibitory potential
• Minimal induction of drug metabolizing enzymes
• Small first-pass effect
• Moderate plasma binding < 90%
• Wide therapeutic index

The implications for DMPK to support drug discovery in the identification of optimal lead candidates are quite valuable. In support of BIOLOGY, DMPK can access in vivo efficacy of compounds in animal models. In support of MED CHEM, lead optimization should be driven by ADME data in addition to pharmacology (HTS) data; early ADME screening is a necessity. The ADME screening advocated here is rational (issue-driven) screening that can be flexible to meet the specific needs of the project. ADME data should be integrated with HTS data as soon as possible and databases should be used for data mining. It is paramount that the screen can resolve good compounds from bad compounds and a simple "rank order" approach is often sufficient. The main objectives of DMPK screening are illustrated below.

At the present time, a rational ADME screening is advocated. This approach requires a careful balance of in vivo studies and in vitro screening (in higher throughput), a careful balance of ADME screening versus prediction, and a flexible screening strategy that can adapt to the needs of the project. However, note that there are several needs for improvement: to increase screening throughput, become proficient at data mining and QSAR analysis, improve in vitro to in vivo correlations, and improve in silico to in vitro correlations

Future

The present status of screening models includes in vitro tests for P450 inhibition and induction, Pgp influx and inhibition, metabolic stability, P450 phenotyping, and plasma protein binding; in vivo screening tests include absorption, plasma half-life, routes of clearance and routes of metabolism. Although none of the systems currently employed for drug metabolism and pharmacokinetic screening can be considered truly high-throughput, many of them are rapid enough to be a practical part of the screening paradigm for modern, fast-moving discovery programs. Challenges faced by Drug Metabolism will be achieving higher throughput, integrating data into meaningful QSAR relationships, and achieving better prediction of ADME profiles and drug interactions in man.

Return to Proceedings »