Development of baseline quantitative structure-activity relationships (QSARs) for the effects of active pharmaceutical ingredients (APIs) to aquatic species

Abstract

The aim of this work was to develop predictive approaches for acute and chronic toxicity in fish, Daphnia, and algae utilizing baseline toxicity models. Currently available public active pharmaceutical ingredient (API) ecotoxicity data were compared to published baseline toxicity QSARs and classification schemes for industrial chemicals. The results showed that methods of assessing ecotoxicity for industrial chemicals are not adequate for the assessment of APIs. To develop equivalent prediction methods for APIs, acute baseline toxicity QSARs for APIs based on hydrophobicity (as log P) were constructed, and the lower limits of toxicity for the public API data were compared with published industrial baseline toxicity QSARs for fish, Daphnia, and algae. These baseline toxicity QSARs were subsequently compared to the available acute toxicity data from the iPiE database. Since 75% of APIs are ionizable, baseline toxicity QSARs were also constructed using log D at pH 7.0. For chronic toxicity baselines, uncensored NOEC and LOEC data from the iPiE database were plotted using either log P or log D at pH 7.0. An alternative methodology was used to develop chronic baseline toxicity QSARs which consisted of iteratively refining the line of best fit until approximately 90% of the values were above the baseline toxicity QSARs. These chronic baseline toxicity QSARs could subsequently be used to identify groups which exhibit toxicity in excess of the baseline (i.e., greater than 10× the hydrophobicity-predicted toxicity).