Workshop Report—KNAPPE (European Union Sixth Framework Project) final meeting

Introduction

The final workshop of the European Union project Knowledge and Need Assessment of Pharmaceutical Products in Environmental Waters (KNAPPE: http://www.knappe-eu.org) was held at the École des Mines, Paris, France, between 8th–9th September, 2008. The meeting attracted over 50 delegates from across Europe with attendees being drawn mainly from the scientific, dispensing, industrial, regulatory, and policy making communities. The two-day workshop provided presentations of results of earlier activities, and round table discussions aimed at formulating proposals for future European actions. These also promoted networking between various end users from different sectors. It aimed to facilitate an exchange of ideas, a review of progress on minimising impacts of pharmaceutical products (PPs) in the aquatic environment in Europe, and included both policy and technical issues. The workshop was tasked with identifying important issues to be tackled by key players, including policy makers, manufacturers, and scientists involved in monitoring PPs in the aquatic environment in the future.

KNAPPE was funded for 18 months and its brief was to look at all aspects of manufacture, distribution, dispensing and end-use of pharmaceuticals for human and veterinary use, and how these factors can impact on the aquatic environment. The project comprised five scientific work packages that reviewed the occurrence of PPs in the aquatic environment, their removal by water treatment processes, methodologies to limit their discharge, their health and environmental impacts, eco-pharmacostewardship and eco-pharmacovigilance. All these aspects were covered in the workshop. The remainder of this report will focus only on the outputs and implications for future monitoring of PPs in the environment.

Challenges in monitoring pharmaceuticals

PPs have been found in surface and ground waters and occasionally in drinking water, and their monitoring poses serious challenges. Currently PPs are not on priority pollutant lists (e.g., Water Framework Directive list of chemicals to be monitored) in Europe. However, they are beginning to appear on various lists of pollutants of special interest at national levels. It is estimated that across Europe over 3000 different products are in use and have the potential to be released into the environment. From desk studies carried out in KNAPPE it was apparent that of these only 200 or so compounds have been investigated in detail in the aquatic environment. The physicochemical properties of this suite of chemicals covers a wide range and this is even wider when their environmental transformation products (e.g., within sewage treatment plants (STPs)) and human and animal metabolites are taken into account. Existing environmental toxicity assays and end points are not necessarily suitable for measuring the impact of PPs. The toxicity of these substances, and hence the associated environmental risk, is very variable and this needs to be considered in targeting monitoring activities. Monitoring is further complicated by the range of sources and pathways by which PPs can enter the environment. These include manufacture and formulation processes, inputs from treated animals and humans, and disposal of unused products. Many PPs are present at only very low levels (pg L⁻¹) but concentrations can vary widely in time and with distance from discharge points (where levels can reach ng to µg L⁻¹). Concentrations in the water column may not reflect total load in a water body since some PPs can binding to sludge and sediments. Analysis of water samples, and the identification of causes of toxicity in water samples is not straightforward since unlike in controlled laboratory experiments compounds are rarely present on their own; rather they are just one pollutant in a complex “chemical soup”. These factors have to be considered in the design of sampling campaigns.

Potential monitoring strategies

Over the last five years there has been considerable progress in the development of laboratory-based methods (typically based on various LC-MS-MS approaches) for the analysis of PPs in various water bodies. Most of these techniques are sensitive and quantitative and can measure the various compounds at the trace levels that are typically found in the environment. However, reliable monitoring of this range of compounds is currently difficult to achieve because of a lack of suitable sampling methods. Most monitoring uses infrequent spot (bottle or grab) water samples and this does not provide a representative picture of levels of PPs where concentrations fluctuate in time due to factors such as variable inputs or weather events. There is an urgent need for more representative sampling methods such as continuous and time- and flow-weighted sampling devices. However, delegates recognised that their wide spread use is limited by considerations such as cost and security. Alternative approaches were discussed, and included the use of passive samplers. Two designs (polar Chemcatcher^® and the POCIS) are available for the sequestration of polar compounds such as PPs. Several field trials in surface water have been conducted with the POCIS. However, the data obtained are of limited value since they are often expressed in terms of ng per sampler rather than absolute concentrations in the water column. Whilst these provide profiles of compounds present they are not useful in supporting risk assessments. In order to obtain measurements of concentration from these data it is necessary to obtain reliable calibration data for the devices for this range of substances. The reliability of the data obtained would be much improved if performance reference compounds (that can be added before use, and whose rate of offloading can be used to account for variations in field conditions during deployment) were available for these types of polar sampler. Participants felt that, based on current experience with non-polar and semi-polar pollutants, passive samplers could offer a more reliable and cost effective approach to monitoring than is possible with spot sampling.

The discussions highlighted a further difficulty associated with the lack of a standard approach to monitoring by the scientific community. Often this is haphazard and data reflect the interests of individual researchers rather than the distribution of PPs. Further, they vary in quality and reliability, and frequently sampling is done in areas (e.g., STPs, discharge points) where high concentrations are expected in order to validate new analytical methods or technologies. These data are presented in a range of formats and even in different units. This increases the difficulties and cost of using this information for purposes of comparison, and in developing models. These data provide a very poor basis for estimating maximum environmental concentrations used in risk assessments, and may lead to inappropriate conclusions and actions.

The reliability of risk assessments depends on not only sound estimates of concentrations in the aquatic environment but also on reliable toxicity data. There is a shortage of the latter, and it is difficult to transfer assay systems from the laboratory, where individual compounds are tested under fixed conditions, to the field situation. In the environment PPs are present as part of a complex cocktail of pollutants from a wide range of chemical classes and sources. It is currently not possible to use direct toxicity assessments for monitoring individual compounds in the environment, and measuring concentrations of individual compounds in organisms is difficult and expensive. Where compounds are present in mixtures, interactions (e.g., antagonism or synergism) can occur. Furthermore, in the field compounds are present along with suspended solids and dissolved organic materials (e.g., humic and fulvic acid) that affect bioavailability. It is difficult to measure bioavailable fractions with classical bottle sampling linked with either sample filtration or direct analysis. Current information based on acute and chronic assays indicates that most pharmaceuticals are relatively non-toxic to aquatic organisms. However, there are some important exceptions (e.g., the endocrine disrupting compounds such as oestrogens), and it is difficult to predict environmental toxicity on the basis of vertebrate data available as part of a product's registration package.

Way forward

The round table discussions highlighted the difficulties faced in monitoring this vast range of chemicals in the various environmental compartments and highlighted the paucity of data currently available. It was agreed that within existing, and projected, monitoring budgets it would be impossible to investigate all PPs, and there is a need for a rationale for identifying those of highest concern. Others could be treated as of much lower priority. Compounds used in very small amounts (e.g., some anti-cancer drugs used in kg per year) could be omitted. A suggestion was to use one compound as a surrogate for a set of others on the basis of similar physicochemical properties or pharmaceutical mode of action, but it was recognised that large uncertainties were associated with this approach. Another approach to reducing monitoring efforts would be to identify a set of sampling sites that are representative of the various types of STPs; though again it was recognised that there can be great variation in efficiency between STPs of similar design but at different locations. A better understanding of the efficiency of different types of STP in removing PPs is needed with a view to designating representative monitoring sites in the future. It was recognised that there is a need to develop appropriate biological assays to support risk assessments for these compounds that exhibit a wide range of biological activities. This would further help to focus monitoring efforts.

It was agreed that currently data from much of the monitoring activity reported in the scientific literature is not suitable for use in risk assessments or in the development of models, and is of limited use to the wider end user community. This could be avoided by the establishment of a central repository for monitoring data. This should use a well defined standard format for both chemical and biological data, and the quality of data should be checked before inclusion. This could then incorporate information from a wide range of sources including routine regulatory monitoring and academic research. The NORMAN (Network of Reference Laboratories for Monitoring Emerging Pollutants; http://www.norman-network.net) project database provides a useful basis for the development of such a repository for monitoring data for PPs. A uniform format would facilitate the development of predictive models (based on e.g., QSAR or LSER), the design of intelligence led monitoring and biological testing. This would reduce animal testing and need for extra water monitoring activity, shorten the time necessary for risk assessments for registration purposes, and thus save money. As data were accumulated in such a database it would be possible to refine existing models. Lastly it was thought that a post-registration scheme for reporting adverse environmental effects of PPs would focus monitoring efforts where needed. However, it is difficult to identify who should fund such a database.

Graham Mills

University of Portsmouth, Portsmouth, UK

Richard Greenwood

University of Portsmouth, Portsmouth, UK

Benoit Roig

École des Mines, Alès, France

---