In August last year, the Environment Agency was wrong-footed by a report in the Observer that the antidepressant Prozac could be found in rivers and drinking water. The report was erroneous; Prozac - or fluoxetine to give the drug its chemical name - has not actually been measured in the UK environment, and is very unlikely to be present in drinking water.
However, fluoxetine has been detected in US and Canadian rivers as a result of discharges of domestic sewage effluents. As more than four tonnes of the drug was prescribed by doctors in the UK last year, it would not be surprising if it was also found in UK rivers.
It fell to the Agency to reassure the public as best it could that if present, levels of the drug were likely to be too low to cause effects. However, the Agency is aware that is the best that it can say. Fluoxetine is a persistent, toxic and highly biologically active compound. It alters the levels of serotonin in the brain and belongs to a class of drugs that have been likened to endocrine disruptors because of their impacts on reproduction in some molluscs.
Many of the thousands of other licensed pharmaceuticals and veterinary medicines may also present unknown environmental hazards.
Emma Pemberton of the Agency's science group told a recent conference hosted by the Society of Chemical Industry: "We simply don't know whether [drugs and drug metabolites] in the environment are going to lead to significant impacts".
Data in the UK have been particularly hard to find, and the Agency has had to rely on data from overseas or back-of-the-envelope calculations to support its contention that, compared with toxic substances like pesticides, drugs are a lower priority (ENDS Report 304, pp 25-28 ).
Dr Pemberton said that there were currently some 5,300 pharmaceutical active ingredients licensed in the UK. The top 50 medicines sold in quantities exceeding 10 tonnes per year include the painkillers aspirin and paracetamol, the anti-epileptic carbamazepine and antibiotics such as amoxicillin and erythromycin.
While human medicines are likely to enter water through the sewerage system, veterinary medicines are likely to end up on farmland soils - possibly leaching into watercourses. Some 450 such medicines are currently licensed, the majority of which are anti-microbials or coccidiostats.
The data on quantities of drugs entering the environment are fragmented, Dr Pemberton said, and it was difficult to get comprehensive information.
However, the Agency and the Environment Department (DEFRA) have commissioned some substantial studies on drugs in the water environment in recent years and have succeeded in filling at least some of the knowledge gaps. Deciding which drugs are the priority and detecting them in environmental samples have proved very challenging (see box ).
Dr Pemberton's conclusions were that levels of both human and veterinary medicines detected in the environment so far are too low to cause acute effects. The Agency's current focus is therefore on the risk of chronic, long-term effects or the potential for interactive effects from the cocktail of substances likely to be present.
Increasingly, the Agency's approach to the drugs issue is being guided through its chemicals strategy published in 2003 (ENDS Report 345, pp 39-40 ). The strategy requires the Agency to admit the limits of its knowledge and acknowledge the issues surrounding many chemical products.
A case in point was the briefing note on fluoxetine the Agency issued in April. It set out the currently available environmental information - the quantities used, excretion and environmental fate - and asked the pharmaceuticals industry to assist in the design and implementation of a monitoring programme.
"The statement was important for us because of the press reports...we wanted to set the record straight," said the Agency's policy advisor Jo Kennedy. "But fluoxetine is like many other compounds which have been on the market for some time. We need the industry to convince us the levels in the environment are safe and to do some chronic [toxicity] tests."
Fluoxetine has been close to the top of the Agency's priority list for environmental monitoring for some time, because of its persistence and toxicity. However, the Agency does not have the capability to detect it in the environment. So far, its research and monitoring efforts have been directed towards other compounds.
The Agency's plea to fluoxetine manufacturers comes on top of a substantial want list published in a position statement on human pharmaceuticals in the environment in 2003.1 The Agency called upon the industry to underpin environmental risk assessment of pharmaceuticals by leading a research programme to examine the potential for chronic environmental effects. It suggested that drugs from all of the most common therapeutic groups should be considered at the kinds of levels being reported in the environment.
The Agency also asked the industry to collaborate with water companies in establishing the fate of drugs during sewage treatment processes, to develop analytical methodologies and to inform the Agency of any low-tonnage chemicals that they might use which were highly toxic or potential endocrine disruptors.
Notably, the paper also expressed the Agency's long-held view that it is time for a management strategy to counter oestrogenic discharges from sewage treatment works.
The Agency fought hard during the water industry's periodic review last year to ensure that a pilot project on removing oestrogens from sewage was funded (ENDS Report 355, pp 3-5 ). One of the most intransigent and active oestrogens to be removed is ethinyl oestradiol, a key ingredient of the birth-control pill.
The results of this study, and emerging evidence that the performance of individual sewage treatment works in removing drugs is highly variable (see box ), may well lead to a new effort to improve the performance of sewage works.
At the SCI conference, Alex Tait of the Veterinary Medicines Directorate contrasted the regulation of veterinary medicines and human drugs. While veterinary products must be assessed for environmental safety before authorisation - and may be refused if unsatisfactory - an EU Directive on human medicines specifically states that authorisation "will not be refused on grounds of environmental safety".
Dr Tait said that the most veterinary medicines used in the UK had now been assessed for environmental safety since a requirement to do so came into force in 1993.
By contrast, his personal view was that "very few if any [human medicines] have been truly assessed for their environmental risks." Although manufacturers do conduct risk assessments for human medicines - as required under article 8(3)(g) of a 2001 EU Directive on human medicines - he suggested that the lack of guidance means that these may not be very vigorous.
Guidelines for the risk assessment for human and veterinary medicines are produced by the European Agency for the Evaluation of Medical Products (EMEA). The guidelines for human medicines were first issued in draft in 1999 and have since been through two more iterations. The latest version - still in draft - was issued for consultation in January.
The Agency has played a major role in shaping the guidance so far and sent a powerful response in calling for the latest draft to be beefed up.
The guidelines propose a tiered assessment starting with a pre-screening stage in which environmental exposure is estimated from likely usage, excretion and degradation rates. If the predicted level of the active ingredient in receiving waters does not exceed an arbitrary threshold of 0.01 micrograms per litre, no further action is required.
The Agency maintains that this threshold should be dropped. Instead it wants to see all products likely to be produced above one tonne per year to proceed to the next stage - phase IIA - of the risk assessment. This would mean that for most products manufacturers would have to produce a basic data set of environmental data, including biodegradation and chronic aquatic toxicology data.
Phase IIA tests should also be brought into line with EU guidelines for the testing of industrial chemicals, the Agency argues, which cover water, sediment and soil compartments and secondary poisoning. Such precautions are justified, it maintains, because of the biologically active nature of pharmaceuticals and the increasingly stringent requirements being introduced for industrial chemicals under REACH.
Metabolites should also be dealt with during phase IIA, the Agency believes, since few products will progress to the next stage - phase IIB. This is only required if the predicted environmental concentration of the active ingredient exceeds any of the predicted no-effect levels for toxicity.
Where the risk assessment indicates potential risks, the Agency wants to see recommendations included in the guidance on risk management. There should be a requirement for post-approval monitoring and consideration should be given to conducting a comparative risk assessment between equivalent products.
Finally, the Agency wants to see a requirement for risk assessments to be revisited when drug usage levels exceed the original risk assessment's assumptions.
"We support initiatives by the pharmaceutical industry on product stewardship and the design of effective alternatives with lower potential for environment impacts," the Agency says in its the position paper on pharmaceuticals. However, it is not clear that any such initiatives actually exist.
No role for the Agency
Another major difference in the licensing of veterinary and human medicines is that the Agency has no access at all to environmental data submitted on human drugs. At least for veterinary medicines, the Agency advises the Veterinary Products Committee and can contribute its expertise on risk assessment and participate in the authorisation decision procedure.
"We would like to see the [pharmaceuticals] data", Ms Kennedy told ENDS. "At the moment there is no formal role for the Agency so we see nothing - which is a contrast not only to veterinary medicines but also to pesticides and biocides."
In its position statement on human pharmaceuticals, the Agency also calls for non-confidential information on pharmaceuticals to be placed in the public domain by EMEA and the Department of Health's Medical and Healthcare Products Regulatory Agency (MHRA).
As far as access to information on drugs is concerned, the Agency is also out in the cold. Its only route for obtaining environmental information on pharmaceuticals is to request it from manufacturers - and there is no obligation for them to provide it.
One avenue which appears as yet untried, is for information to be requested on drugs under the Freedom of Information Act 2000. Key physical and chemical data, including toxicity and biodegradation could well be within the scope of this legislation.
On fluoxetine, the Agency called for the pharmaceutical industry to help in the design and implementation of a monitoring survey to measure levels of the drug and its metabolite in UK waters. It also wanted help to assess whether the levels found are likely to have any environmental impacts.
Ms Kennedy said that the Agency has not yet had any response. The drug was first marketed by Eli Lilly but the patent ran out in 2000 and some 19 other companies now sell fluoxetine in the UK.
"We consulted with Eli Lilly during the drafting of the briefing and would expect they are the group who will be most willing and probably best placed to follow up with us on the actions we are calling for," Ms Kennedy added. "They could justifiably argue, though, that responsibility should be shared with all the other companies now marketing fluoxetine."
Generic drug manufacturers now supply some 99% of the UK market for fluoxetine and are represented by the British Generic Manufacturers Association. But this body has yet to respond to the Agency's attempts to open a dialogue on pharmaceuticals in the environment, Ms Kennedy said, and it also refused to return calls from ENDS.
But the Agency has at least had some contact with the major players in the industry. Following its 2003 position statement, it has met with companies like Pfizer, GlaxoSmithKline and AstraZeneca. There have been some developments, Ms Kennedy says, but it has been slow - particularly because many drugs are now generic and the Agency needs to deal with many manufacturers or importers.
Mike Murray, head of manufacturing and environment at the Association of the British Pharmaceutical Industry, commented: "We feel the Agency has a legitimate interest but there is no evidence of any problems in the environment, so it has to be kept in perspective."
The Association represents many of the major players who are likely to be developing new products and is "highly supportive" of the risk assessment guidance proposed by the EMEA, Mr Murray says. However, there is little sign that the Agency is going to get the kind of help it has asked for on data and risk assessment of older products. Such information is hard to come by, Mr Murray concedes, and may well not exist.
Without legislative action to require the pharmaceuticals industry to risk-assess older drugs, it is likely that regulators like the Agency will have to rely on their own research.
While the EMEA guidance on risk assessment is still in draft there is little obligation to conduct risk assessment even on new products. Nevertheless, many companies were following the procedure at least in outline, Mr Murray said.
EA vs. EMEA?
Once risk assessments are finally required to support registrations, the industry favours the EMEA as a central agency providing in-house expertise to review and assess them, Mr Murray said. He showed little enthusiasm for relying on external input from organisations like the Environment Agency: "Most new registrations now go through a central agency and that should set the template for how they are processed."
The EMEA says that discussion of the comments received on the consultation is ongoing and, depending on the outcome, the final guidelines "may be released by the end of the year".
Data on environmental risks submitted to the EMEA in application dossiers are "assessed by experts from regulatory agencies for medicinal products in the EU Member States, environmental agencies and/or independent external experts," the EMEA says.
However, the Environment Agency says it has yet to be consulted by the MHRA on any risk assessments, even though it has sought involvement.
"We wrote to them earlier this year saying that we had a specialist team which is very knowledgeable in this area, but we have yet to hear back," Ms Kennedy told ENDS.