US FDA's regulations in 21 CRF part 25 specify that environmental assessments (EA) must be submitted as part of new and abbreviated new drug applications, biologics licence applications, and for various other actions except when a claim for categorical exclusion is applicable. The final rule (62 FR 40569) became effective on August 28, 1997. A guidance for industry was issued in July 1998 in order to provide information on when an EA should be submitted and on how to prepare EAs for submission to CDER and CBER).

EU Council Directive 2001/83/EC, as amended, states that applications for marketing authorisation of a medicinal product for human use shall be accompanied by an environmental risk assessment (ERA). A draft guideline providing guidance on the data requirements and principles for an ERA was developed by EMEA CHMP. The latest version was issued on January 20, 2005. At the EMEA conference held in London in October 2005, participants from interested parties (EU and national competent authorities, academia and industry) met to discuss key issues and to clarify the legal requirements of the new legislation. Finalisation of the guideline could be possible early in 2006, depending on the number of necessary changes (Source: Press Release, Doc. Ref. EMEA/359945/2005).

General assessment principles

The assessment principles are comparable in the US and EU guidance documents. Potential risks to the environment are evaluated in a tiered procedure, which can be terminated when sufficient information is available, either to indicate the medicinal product is unlikely to represent a risk to the environment or to characterise it sufficiently.
The focus is on the aquatic environment assessing the risk after excretion of the medicinal active substance (and/or its degradation products) by patients and subsequent transport via sewage treatment facilities to natural waters. Terrestrial exposure is assumed if the compound is strongly bound to sewage sludge which is landspread. In both documents, limits of "acceptable" concentrations in surface water (PEC or EIC)4 have been established above which an environmental assessment is required. The following "action limits" apply:

. US EICaquatic5 =1µg/l, reached with sales of 50 tons active moiety per year
. EU PECsurface water =0.01 µg/l, reached with a prescribed dose of 2 mg per day/patient

They are based on generic assumptions about the amount of drug emitted by patient populations to sewage treatment works and finally, to surface waters. Refinement of the estimates is possible in higher tiers of the assessment, taking into account information from pharmacokinetic/toxicokinetic studies, abiotic/biotic degradation and dilution processes and actual market penetration. The PEC/EIC values are compared with effect concentrations from single species tests by applying assessment factors to account for the uncertainty of extrapolation from data with a limited number of test species to the "real environment". The resulting PEC/PNEC ratios are the basis for the decision as to whether the risk is considered acceptable, further testing is triggered and/or risk mitigation measures are applicable.

Data requirements for environmental risk assessments

The data requirements for physico-chemical properties are comparable. Results from environmental fate studies (e.g., hydrolysis, photolysis and ready biodegradability tests) should be submitted, if available, in order to provide information on potential abiotic and biotic degradation processes. Usually, the data need not be generated solely for ERA purposes but can be derived from information developed for the physicochemical characterisation of the drug. There are differences between the EU and US in the data required for aquatic ecotoxicity. For US Tier I assessments, acute testing on at least one representative species from an aquatic base set is required. In the EMEA/ CHMP guideline, continuous emissions of the medicinal substance and exposure of the aquatic environment is assumed. Therefore, instead of acute toxicity testing, subchronic tests with fish and aquatic invertebrates are required in Phase II, Tier A. However, submission of the subchronic test package generated for the EU is expected to be acceptable to CDER/CBER as a higher tier surrogate for acute testing, in particular, since the requirements for testing on the toxicity to algae are the same. Therefore, the data package generated for EU submissions should support applications under both sets of legislation, provided the tests are done according to internationally accepted test guidelines and GLP standards. Compilation of "combined" test programme in order to meet the data requirements of both regions is also possible for higher tier testing.


View Table Assessing environmental risk of pharmaceuticals
Table 1: Test programme for EU Phase II, Tier A (EU) and Tier I (US)
*) other test guidelines, e.g., OPPTS may be acceptable but need to be confirmed


View Table Test programme for EU Phase II, Tier B and Tier II (US)

Table 2: Test programme for EU Phase II, Tier B and Tier II (US)

Performing right tests

According to the EMEA/CHMP draft guideline, the Koc can be determined using an HPLC method (OECD 121), or the batch equilibrium methods using soils (OECD 106) or sludge (OPPTS 835.1110). The use of the method has to be justified. The methods differ significantly in their test design, detail and quality of information generated, although they finally produce the "same" endpoint (Koc). Without further guidance, a pragmatic approach is to check whether the HPLC method is applicable. Otherwise, it is recommended to conduct one of the batch equilibrium methods, depending on the compartments likely to be exposed and later use of the endpoint in the assessment. A ready biodegradability test is only mandatory in Phase II, Tier B. However, it may be useful to conduct such a study at an early stage. In the event that the substance is readily biodegraded to carbon dioxide (e.g., in OECD 301 B), waiving of the "water/ sediment" study (OECD 308) might be possible.

On the other hand, ready biodegradability tests give only very limited information on the fate of pharmaceuticals in surface waters and sewage treatment systems, where the substance is not ultimately degraded to CO2. In any case, more information on the fate of pharmaceuticals in aquatic systems can be obtained from sewage treatment simulation tests (OECD 303 A, Figure 6) and/or from a water/sediment study (OECD 308). These tests provide comprehensive information on the depletion processes involved, including degradation to metabolites and adsorption to sludge or sediment.

For example, the elimination of three pharmaceutical substances found to be non-toxic to sewage micro-organisms and not readily or inherently biodegradable in respective tests was compared to the results derived from sewage treatment simulation tests and water/sediment studies.

For diazepam, no degradation was found in the MITI-II and the simulated sewage treatment tests, whereas the OECD 308 (water/ sediment) showed a fast dissipation from the water phase mainly by binding to the sediment. This shows that really detailed information on the various processes involved could finally only be obtained from aquatic transformation tests done according to OECD 308 test guidelines (water/sediment study under aerobic conditions).

Ibandronate showed a very limited degradation in the "bottle tests" and the simulated sewage treatment test resulted in not very consistent data. Again here, the OECD 308 provided the most reliable data and gave the most comprehensive answers. In the water/ sediment study, ibandronate disappeared extremely rapidly from the water phase by binding very tightly to the sediment, which was not necessarily expected when considering its high water solubility and low estimated Koc (based on logPow/QSAR). No formation of CO was observed which confirmed the results of the biodegradation tests in bottles. This demonstrates that for ionic aliphatic substances like ibandronate, the Koc estimated by QSAR is not a reliable indicator of the adsorption behaviour.

Fluorouracil showed no degradation in the "bottle tests" but was completely eliminated in the simulated sewage treatment test, i.e., mineralised to CO2. In this case and at environmentally relevant concentrations, 14C-labelled fluorouracil was rapidly degraded to CO2.

In general, the comparison shows that depending on the predicted or known substance properties, a different and tailored testing strategy may be suitable for comprehensive assessment of the potential of pharmaceuticals to be removed in sewage treatment facilities and their fate in surface waters. Some flexibility in the test design and sequence of testing should be allowed. The results of this comparative investigation were presented in a poster at the SETAC conference held in Lille, France (22-26 May 2005).

View Table Physico-chemical properties and biodegradability
Physico-chemical properties and biodegradability

(The authors, Enrico Kiefer is Project Leader / Head of Product Safety, Business Unit Agro and Uwe Morgenroth, Operational Unit Manager Environmental Safety and Metabolism, RCC Ltd.)