Prioritization of pharmaceuticals in drinking water exposure based on toxicity and environmental fate assessment by in silico tools: An integrated and transparent ranking

Abstract

The concentration of pharmaceuticals in water and health risk to the human population is a topic of concern among the scientific community, regulators, and the general population. Many recent studies have suggested different, even conflicting, criteria for the definition of priority pharmaceuticals for drinking-water exposure. In this work, we assessed 10 endpoints of environmental fate and human toxicity for 39 active pharmaceutical ingredients (APIs) by in silico models and data review. A visual and transparent Toxicological Prioritization Index (ToxPi) was proposed based on in silico, in chemico, in vitro, and in vivo results. Predictions of environmental fate and toxicity were made through in silico models of EPISuiteTM, PBT Profiler v.2, TEST v4.2 and QSAR Toolbox v.3.2. The software ToxPi GUI was used to create a multicriteria-based ranking with environmental fate and human toxicity endpoints to identify priority compounds. Our multicriteria ranking (fate and hazard-based) suggested that the some of the most concerning APIs were metronidazole (ToxPi: 5.67; CI95%: 5.51–5.83), nitrofurantoin (ToxPi: 5.22 CI95%: 5.21–5.22) and ethinylestradiol (ToxPi: 5.20; CI95%: 4.70–5.79). Different APIs may represent high level of concern by different characteristics of environmental exposure and/or toxicity, and our visual muticriteria ranking can support targeted decision-making. Computational tools may be important for predicting environmental fate and toxicity as well as visually identifying priority compounds by multicriteria-based rankings and predicting compounds with high priority to support more targeted multicriteria decision-making, risk assessment, and monitoring steps.