Test cascade

The key to successful lead identification and optimization is the test cascade. This is a series of hierarchically arranged tests or screens (usually biological) which allow the progressive optimization of chemical structure, via the iterative 'making and testing' approach, to give the desired performance. In industry, the desired performance will normally be defined by a target profile listing the essential and desirable features of the prospective drug. This will define the mechanism of action and biological effect, as well as provide measures of physico-chemical properties, potency, selectivity, pharmacokinetics, and therapeutic index. A test cascade must be robust, rapid, and efficient.

The primary screen is usually a biochemical test using the recombinant protein target or a genetically engineered cell line. A selectivity screen will usually follow, together with an assay to confirm the biological effect or a surrogate thereof in intact cells. Ideally, an assay should be available to confirm that this effect is achieved by the desired molecular mode of action. Activity against tumour cells in culture is normally sought.

The transition from activity in cell culture to pharmacological effect in the intact animal is a challenging one. The latter requires the compound to display robust, 'drug-like' character, and inadequate pharmacokinetic/ADME (absorption, distribution, metabolism, and excretion) properties are often limiting. Hence a screen for blood levels, often using mixtures of compounds administered in cassette or cocktail dosing for higher throughput, can be useful.

Currently under evaluation by the NCI and others is the hollow fibre assay that has elements of an in vitro-in vivo hybrid test and gives an indication of the achievement of concentrations in vivo sufficient for at least minimal activity against human tumour cells.

The final stage of testing will involve seeking evidence of regression or growth arrest/delay in a human tumour xenograft. Depending on the biological effect sought, more complex tests such as orthoptic or


Number of compounds tested

Biochemical screen (Usually high throughput test with recombinant reagents)

■ Rational design

■ Structural biology

■ Combinatorial chemistry

Target cells assays

(To measure cell activity, selectivity, mode of action)

Typically 200 000-500 000

■ Rational design

■ Structural biology

■ Combinatorial chemistry

1000-5000 (Typically 0.5-1% hit rate from high throughput screen)

Pharmacokinetic endpoint (e.g. blood level as a measure of drug exposure)

and/or Surrogate endpoint (e.g. normal tissue or animal tumour response as a measure of biological effect)

Disease model (e.g. human tumour xenograft)

Specific organ toxicity 50

Structure of a contemporary mechanism-based drug discovery test cascade.



metastatic models may be useful. Transgenic mouse models can be valuable, as can surrogate non-tumour endpoints.

Usually, activity will be sought in at least a small panel of tumours, including human xenografts. Ideally, these would be characterized for the molecular target and pathway involved, together with any other relevant features. Such a panel can be especially useful in selecting from a shortlist of potential clinical development candidates that may all display the target profile. An alternative to this scheme is a modification to the NCI in vitro cell panel approach, whereby compounds are selected based on correlation of sensitivity with molecular parameters.

0 0

Post a comment