Environmental risk assessment for the serotonin re-uptake inhibitor fluoxetine: Case study using the European risk...

The serotonin re-uptake inhibitor fluoxetine was selected for an environmental risk assessment, using the most recent European guideline (EMEA 2006) within the European Union (EU)-funded Environmental Risk Assessment of Pharmaceuticals (ERAPharm) project due to its environmental persistence, acute toxicity to nontarget organisms, and unique pharmacokinetics associated with a readily ionizable compound. As a widely prescribed psychotropic drug, fluoxetine is frequently detected in surface waters adjacent to urban areas because municipal wastewater effluents are the primary route of entry to aquatic environments. In Phase I of the assessment, the initial predicted environmental concentration of fluoxetine in surface water (initial PEC(SW)) reached or exceeded the action limit of 10 ng/L, when using both a default market penetration factor and prescription data for Sweden, Germany, and the United Kingdom. Consequently, a Phase II risk assessment was conducted in which green algae were identified as the most sensitive species with a NOEC of <0.6 microg/L. From this value, a predicted no effect concentration for surface waters (PNEC(SW)) of 0.012 microg/L was derived. The PEC/PNEC ratio was above the trigger value of 1 in worst-case exposure scenarios indicating a potential risk to the aquatic compartment. Similarly, risks of fluoxetine for sediment-dwelling organisms could not be excluded. No risk assessment was conducted for the terrestrial compartment due to a lack of data on effects of fluoxetine on soil organisms. The need for a separate risk assessment for the main metabolite of fluoxetine, norfluoxetine, was not conducted because of a lack of fate and effect studies. Based on published data, fluoxetine and norfluoxetine appeared to have a low to moderate bioaccumulation potential, which should be confirmed in formal studies according to OECD guidelines. Exposure assessments for fluoxetine according to the current framework rely heavily on K(OC) and K(OW) values. This approach is problematic, because fluoxetine is predominantly a cationic substance at environmental pH values. Consequently, the fate of fluoxetine (and other ionic substances) cannot be predicted using partition coefficients established for nonionic compounds. Further, published estimates for partition coefficients of fluoxetine vary, resulting in considerable uncertainties in both the exposure and environmental risk assessments of fluoxetine.