Active Pharmaceutical Ingredients (APIs) from human medicinal products are detected in surface waters globally, but the extent of their impact remains insufficiently understood. Conducting environmental risk assessments (ERAs) including experimental studies for all pharmaceuticals is unfeasible, particularly for legacy APIs approved before the ERA became a mandatory requirement in the European Union. This challenge can be addressed by prioritising legacy APIs for ERA, as highlighted by the European Commission's recent proposal for a revised EU pharmaceutical legislation, calling for scientific criteria to prioritise legacy APIs, using a risk-based approach. In line with this, we present a risk-based prioritisation framework to identify high-priority legacy APIs for further testing and ERA. The framework, which is intended as a screening and prioritisation tool and not as a substitute for regulatory-compliant ERAs, was applied to a dataset of 1,687 APIs. Environmental concentrations in surface waters were estimated using European volume sales data from 2019 to 2023. Effect concentrations were predicted using multiple approaches, including Quantitative Structure-Activity Relationship (QSAR) models, and theoretical Therapeutic Water Concentrations (TWCs). By comparing predicted values against empirical ecotoxicity data, we showed that QSAR models, which generally show acceptable statistical performance, tend to underestimate the effects of highly toxic pharmaceuticals. In contrast, TWC values were more conservative but exhibited greater variability in performance. Exposure and effect concentrations were combined into preliminary risk quotients (pRQs), which were ranked and subsequently benchmarked against a reference ranking for validation. In the final step, multiple pRQs were combined into a single risk-based ranking. This list, along with a summary of ERA data availability for the included APIs, serves as a starting point for stakeholders, such as industry and regulators, to screen legacy APIs for data gaps, guiding decisions on which substances require a further ERA, possibly involving testing, and what types of studies are needed.
Cannata et al. (Mon,) studied this question.