Using an Innovative Design in Behavioural Pharmacology Studies Saves Money and Animal Lives
Anders Viberg (1), John Martino (2), Etienne Lessard (2) and Jennifer Laird (2)
(1) AstraZeneca R&D Södertälje, (2) AstraZeneca R&D Montreal
Objectives: Results from pre-clinical efficacy studies are of importance for many decisions in drug development. Being used to estimate active concentrations in the biophase in man, these results have a direct impact on the evaluation of toxicology and Phase I results and they are also an aid in dose setting in Phase II studies. Typically, pre-clinical studies in analgesia consist of at least three different studies; dose-finding, effect-duration and tolerance development studies. In typical behavioural models, the exposure is measured in a parallel group of animals, which may compromise the precision in describing PKPD relationships. The objective of this study was to investigate if pre-clinical analgesic studies in rats could be more effectively performed using sparse PK sampling in the PD tested animals and thereafter evaluate the results using a population approach.
Methods: A refined dosing strategy was developed and applied for drug X in the rat Chung heat hyperalgesia model. PD measurements were done on day 1, 3 and 5. Two PK samples per day were taken in day 2 and 4. In a separate group PD measurements were done on rats without PK samples taken. Data was analyzed using a population approach in NONMEM.
Results: The animals with PK sampling had the same therapeutic response as the animals without PK sampling. A direct concentration-effect relationship with good precision could be established and no tolerance development was observed. When comparing the new design with to the old design, substantial savings was done. The number of animals was reduced with 44% and the number of working hours in the lab was reduced with 63%.
Conclusions: The new suggested design makes substantial savings in both animal lives and money. Moreover, the exposure response relationship was described with higher statistical precision compared to using the old design.
