Preclinical drug development

This domain generally comprises the screening of substance libraries for eligible candidates, the determination of their activity and specificity and tests of their toxicity and mutagenicity. These experimental studies are largely performed in vitro (cell culture, cell or tissue preparations) and completed by in silico investigations (simulations). The next evaluation steps investigate the fate of the substance in animals (pharmaco-kinetics, PK) and its efficiency in disease models (pharmaco-dynamics).

Our focus is on the key parameters of preclinical PK which comprise absorption, distribution, metabolism and elimination (ADME). We provide specialized methods, equipment and expertise centred on small animal imaging (PET). For suitable candidates, this miniature approach can be scaled up to the application in humans.

Absorption, distribution, metabolism and elimination (ADME)

In some cases these parameters can be obtained using a non-radioactive (cold) test compound and modern mass spectrometric methods. A classical approach uses [14C] radio-labelling of the compounds of interest. Some questions regarding preclinical PK can also be answered using candidate compounds labelled with short-lived (PET) radioisotopes such as [11C] or [18F]. Radioactive test compounds enable a full substance balance at very low detection limits and an easy identification of metabolites. Bio-distribution is directly visualised by exposition of photographic films or digital storage foils (autoradiography). The candidate substance can be applied in vivo or ex vivo as well as post mortem at tissue sections which are immersed in a solution of the candidate substance (radioligand binding study).

Fig. 1: Autoradiographs of the radioligand [3H]CPFPX in rat brain sections. The binding density (fmol/mg protein) is pseudo-colour coded and reflects the different local densities of receptors (example: A1 adenosine receptor).

Fig. 2: Chromatographic separation of extracts from different organs and body fluids of the rat after application of the drug candidate [18F]CPFPX.

Fig. 3: Determination of the chemical structure of the major metabolite M1 of CPFPX starting from the mono-isotopic mass of the molecular ion and masses of fragment ions after collision-induced decay using a triple quadrupole instrument.

Dose finding

With regard to neuropsychopharmaceuticals dose finding is highly challenging as the pharmacodynamic action can build up with a long delay (several weeks). Furthermore, animal models reflect only partially and incompletely the human condition - if any animal model is available. Thus, receptor occupation and displacements studies are applied where brain sections are incubated with an established reporter radioligand in the presence of different doses of a cold test compound. Typically, a dose causing about 70% receptor occupation is considered pharmacologically active. This type of dose finding study can be performed in small animals and be scaled up to the application in humans in clinical phase IIa studies.

Finding of clinical indications / determination of clinical activity

The activity of a drug candidate is usually tested in suitable disease models which mimic the human condition as closely as possible. Beneath cell culture tests animal models of disease are an important tool for drug development. In particular, repeated and/or longitudinal measurements (drug vs. placebo; baseline vs. application) can help to identify treatment effects at suitable target structures (e.g. dopaminergic and adenosinergic systems in Huntington’s disease).

Fig. 4: Use of radioligand techniques in the characterisation of animal models. Differential analyses of dopaminergic receptors and transporters in a transgenic mouse model of Huntington’s disease vs. wild-type ([3H]SCH23390 for D1-, [3H]raclopride for D2-receptors; [3H]mazindole for the membrane bound dopamine transporter).

Fig. 5: Use of radioligand techniques in the characterisation of animal models. Differential analyses of adenosinergic receptors in a transgenic mouse model of Huntington’s disease vs. wild-type ([3H]CPFPX for A1- and [3H]ZM241385 for A2A adenosine receptors).

Autoradiographic studies at human post mortem tissue

Radioligand binding studies, as a subtype of biodistribution studies, are directly performed on fresh frozen human post mortem tissue. These studies allow investigating analogies and differences with regard to human and non-human target structures of a give drug candidate and help to stratify human phase I studies. They also complement phase I studies with high resolution data (100 – 200 µm for [3H] in contrast to 3 mm for in vivo PET).

Fig. 6: Comparative studies of the in vivo binding of [18F]CPFPX in PET and of post mortem [3H]CPFPX at human brain sections (orientated as indicated in the insert). Note high binding in thalamus and cortex and low binding in white matter.