Plasma Concentrations of Human Pharmaceuticals as Predictors of Pharmacological Responses in Fish

Abstract

Pharmaceuticals typically have specific enzyme and receptor-based modes of action, which are extensively studied in mammals during drug development. Based on the considerable evidence of enzyme/receptor conservation across species, a predictive model has been developed that links therapeutic mammalian plasma levels to pharmacological responses in fish. In this model, a measured human therapeutic plasma concentration (H_TPC) is compared to a predicted steady state plasma concentration (F_ssPC) in fish, which results in an effect ratio (ER = H_TPC/F_ssPC) being computed. The lower the ER, the greater the potential for a pharmacological response in fish. The model was applied to twenty-eight drugs representing 15 therapeutic classes. ER values ranged from ≤ 1 to ≥ 10 000, with the category that represented an ER ≥ 10 000 containing the largest number of compounds (12 of 28). The two compounds with an ER ≤ 1 (17β-oestradiol and 17α-ethinyloestradiol) have previously been identified in chronic ecotoxicity evaluations as being active in fish. To begin to validate this model, rainbow trout were exposed for 96 h to 900 ng ml⁻¹ ibuprofen or 200 ng ml⁻¹ carbamazepine. These exposures resulted in measurable levels of these two compounds in trout plasma. These data indicate that pharmaceuticals may partition from water into fish plasma. While additional model refinement is needed, the model can provide a framework and prioritisation tool for considering pharmaceutical responses in fish.