Reinhard Länge

Environmental risk assessment of pharmaceutical drug substances--conceptual considerations.

Drugs, i.e. active ingredients of human medicinal products, may be introduced into the environment after use in patients by sewage effluent pathways and consequently are detected at low concentrations in sewage effluents and in surface waters. Legal requirements in a number of geographical regions (Europe, US, and intended in Canada) demand environmental risk assessments (ERA) for new drug substances. Existing regulatory concepts of ERA are based initially on a set of short-term ecotoxicological studies in three to four different species, environmental behavior and the application of assessment factors to correct for the ERA inherent uncertainty. Based on theoretical considerations and the experience with a very limited, but well investigated, number of examples while considering that drugs are highly biologically active compounds, the appropriateness of this risk assessment procedure for all drug substances might be questioned. Indeed, e.g. long-term effects may occur at much lower concentrations and follow different toxicodynamic mechanism than extrapolated from short-term studies., In such cases, the application of assessment factors for deriving chronic no-observed effect concentration (NOECs) appears to be problematic. Although long-term tests with a variety of organisms would provide a complete database for the evaluation of the environmental risks, this is unachievable for all drugs due to time, money and animal welfare constraints. In order to avoid unnecessary testing, a concept is presented, which makes use of pharmacological and toxicological, as well as pharmaco- and toxicokinetic information derived from mammals during drug substance development. Useful data for adoption in a case-by-case testing strategy can be obtained by evaluating (a) the pharmacological activity, which indicates specific targets in mammalian species and may allow for an analysis, whether a similar target is available in aquatic species; (b) the mammalian toxicity, which may indicate, which targets are most susceptible to adverse effects; (c) the difference between acute and chronic effects in mammals, since the magnitude of this difference may indicate, whether long-term effects are expected at significantly lower levels than acute effects; (d) the (pharmacologically and toxicologically) effective plasma levels in mammalian test organisms, which may be compared with the relevant exposure scenario for the environment. Additionally, activity classes of compounds may be established based on experience with specific substances, in order to develop an appropriate test strategy. The above preliminary considerations should support decisions on the selection of candidate substances for chronic effects studies and for the appropriate selection of test species and endpoints to monitor. Generally, ecologically relevant endpoints such as impairment of growth, development and reproduction should be used to assess the ecotoxicologic effects.

Analysis of the ECETOC aquatic toxicity (EAT) database. II - Comparison of acute to chronic ratios for various aquatic organisms and chemical substances

Environmental risk assessment often requires prediction of potential chronic effects in aquatic species from acute toxicity data. The scientific rationale for using extrapolation factors for this continues to be debated. As a contribution to the debate this paper analyses the acute-chronic-ratios (ACRs) based on acute EC50s (or LC50s) and (sub)chronic NOEC values using information in the ECETOC Aquatic Toxicity (EAT) data base. The ACRs were calculated separately for all aquatic species available and the results presented according to specific substance classes. It was unusual to find more than three species giving ACRs for a single chemical but eight of the 28 species in the analysis gave ACRs for five or more chemicals. Based on the commonly used 90%-ile ACR (whose usefulness is underlined by the data presented here), 90%-ile ACRs were observed from 192 (for metals/organometals to 20 for ‘other inorganics’. For organic chemicals resembling those which may be submitted for registration under the European Community 7th Amendment Directive (92/32/EEC), the results were further analysed based on time-specific ACRs and gave the 90%-ile as 24.5: the maximum ACR observed for such organic substances was 28.3. ACRs for individual taxonomic groups (e.g. fish, daphnids) were also calculated. Except for three cases an ACR derived from studies on Pimephales promelas was within a factor of 2 of the ACRs of five other fish species. In contrast, the ACRs for Daphnia magna varied to a greater extent from the ACRs for other invertebrates. For a quarter of cases the ratios differed by 5.0 or more. The results are discussed with respect to current environmental risk assessment procedures.

Analysis of the Ecetoc Aquatic Toxicity (EAT) database I — General introduction

In assessing the environmental risk of a chemical it is necessary to predict safe levels in the environment, but this requires extrapolation from relatively small databases. To improve understanding of the use of existing data for such extrapolations, the European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) established a Task Force to assemble a high-quality database on aquatic toxicity from published papers. The database contains over 2200 records of effects of about 360 substances on 121 aquatic species. Computer programs were developed to enter, search for, retrieve and analyse the data. Output consisted of simple counts, frequency distributions and regression analyses by which (for example) comparisons of different life-stages, species-sensitivity and acute:chronic ratios, and application of QSARs could be examined for use in risk assessment. This paper describes the content of the database, the methods of analysis and gives a brief comparison of acute: chronic ratios: four subsequent papers describe the results.