Pharmaceuticals in the environment: good practice in predicting acute ecotoxicological effects

Area of Science:

• Environmental Chemistry
• Ecotoxicology
• Computational Toxicology

Background:

• Growing detection of pharmaceuticals in the environment raises concerns about adverse effects on non-target organisms.
• Limited experimental ecotoxicity data for pharmaceuticals necessitates the use of in silico predictive tools.
• Existing models like ECOSAR, developed for industrial chemicals, may have limitations when applied to pharmaceuticals.

Purpose of the Study:

• To assess the reliability of the ECOSAR program for predicting the aquatic toxicity of 364 diverse pharmaceuticals.
• To evaluate the confidence in ECOSAR predictions based on the representativeness of assigned chemical classes.
• To investigate the correlation between predicted and experimentally determined toxicity values for pharmaceuticals.

Main Methods:

• Utilized the ECOSAR program to classify 364 pharmaceuticals into chemical classes and predict their aquatic toxicity.
• Assessed prediction confidence by evaluating the appropriateness of the assigned chemical classes for each pharmaceutical.
• Correlated ECOSAR-predicted toxicity values with available experimental ecotoxicity data.

Main Results:

• ECOSAR provided reasonable aquatic toxicity predictions for pharmaceuticals within its established applicability domain.
• Predictions were less reliable for pharmaceuticals falling outside the model's domain.
• Highlighted the importance of expert judgment in determining model suitability for specific chemical compounds.

Conclusions:

• The ECOSAR model can be a useful tool for predicting pharmaceutical aquatic toxicity, but its applicability domain must be carefully considered.
• Expert evaluation is essential to ensure the appropriate use of in silico models for environmental risk assessment of pharmaceuticals.
• Further research may be needed to refine predictive models for broader pharmaceutical applications.