The relevance of preclinical models to the treatment of postmenopausal breast cancer.

The predictive value of test results in animals when selecting a compound for potential therapeutic human use depends upon the relevance of the animal model to the human disease and the comparative pharmacokinetics of the compound in animals and man. The development of the aromatase inhibitor, anastrozole (Arimidex), illustrates the importance of these factors. In postmenopausal women with breast cancer, aromatase activity in the peripheral tissues is the main source of oestrogen for tumour growth. Only one form of the human enzyme is known, which is not subject to strong feedback control. Inhibition of aromatase therefore simply reduces oestrogen production. This situation is mimicked by assays of acute inhibition of ovulation in rats, chronic inhibition of androstenedione-induced uterine hypertrophy in sexually immature rats, and chronic inhibition of peripheral aromatase in monkeys. In all these assays, maximum anastrozole activity was consistently achieved at an oral dose of about 0.1 mg/kg, and the clearance half-life of 7-16 h indicated that once-daily dosing would be possible in humans. The clearance half-life in postmenopausal women is about 50 h, and with once-daily dosing the dose of anastrozole required for maximal inhibition is 1 mg/day. The rat 7,12-dimethylbenzanthracene tumour model, in contrast, is supported by ovarian oestrogen and chronic inhibition provokes positive feedback loops that try to restore oestrogen production, masculinise the animals and decrease the clearance half-life of anastrozole. Higher doses (10 mg/kg) of anastrozole are therefore needed. Variations in the dose of aromatase inhibitor required in different models, therefore, can be explained in terms of pharmacokinetics and do not reflect the effectiveness of anastrozole as an aromatase inhibitor.