Analytical methods for nitrofurans: Lessons to be learned and new developments.

Glenn Kennedy, Belfast, Northern Ireland, UK

1. Residue Controls in the European Union.

The European Union (EU) has put in place a series of measures to protect the health of the consumer and to ensure that food of animal origin produced within the EU and produced in 3^rd countries to be exported into the EU is safe to eat. Licensed veterinary medicines are evaluated by the European Medicines Evaluation Agency. Based on their evaluation, a Maximum Residue Limit (MRL) is calculated for each medicine in the tissues of each animal species in which the product may be used. Large safety margins are built into the calculation of MRLs. This means that a violative concentration (> MRL) in tissue does not necessarily pose a risk to human health. It does, however, mean that the product is not suitable for the purposes of international trade. All MRLs are listed in the Annexes to Council Regulation 2377/90^[44]. Compounds listed in Annexe I have had a final MRL established (e.g. sulphonamides). Compounds listed in Annexe II require no MRL (e.g. vitamin A). Compounds that are still under evaluation are granted a temporary MRL and are listed in Annexe HI. A further Annexe to this regulation (Annexe IV) lists those compounds for which no MRL can be set. These compounds cannot be used in food-producing animals. The nitrofurans are included in this group.

2. Use of nitrofuran drugs in the EU.

The nitrofurans are a group of antibacterial compounds. The main members of this group are furazolidone. Furaltadone, nitrofurantoin and nitrofurazone (Figure 1). They are relatively broad-spectrum bactericidal drugs. They are active against Salmonella spp., coliforms, Mycoplasma spp., Coccidia spp. and some other protozoa. The use of all nitrofurans, with the exception of furazolidone was banned in the EU in 1993^[45]. The ban was introduced because of concerns over the carcinogenicity of these compounds. Two years later, the ban was extended to cover furazolidone. Again, the reasons for the ban were the carcinogenicity of the parent drug, the extensive metabolism of furazolidone and the lack of information concerning the safety of its metabolites^[46]. Since then, it has been forbidden to use any nitrofuran in any food-producing animal within the EU, or in any animal destined for export into the EU.

3. Monitoring for veterinary drug residues.

The EU requires Member States (MS) to establish National Residues Surveillance programmes^[47] to monitor food of animal origin for the presence of licensed medicines, Annexe IV compounds and for residues of the hormonal growth promoters and ß-agonists, which are also banned in the EU^[48]. The EU requires MS to submit monitoring plans to the Commission for approval every year, and required MS to report results back to the Commission on an annual basis. The Commission has also established a Rapid Alert system to inform MS when potentially harmful residues are detected in products imported from 3^rd countries. This system is designed to speed up MS responses when such residues are detected. All EU MS are therefore required to test domestic product for the presence of nitrofurans. This requirement has been in place since before the compounds were banned in the 1990s. When furazolidone was licensed in the EU. it had a provisional MRL (it was included at that time in Annexe III of Council Regulation 2377/90) of 5 µg/kg. The marker residue was defined as "any compound containing an intact 5-nitro group". In practice, most MS developed analytical methods to detect the parent drug, usually based (at that time) on HPLC with UV detection.

4. Detection of nitrofurans

In the late 1980s, it was shown that furazolidone was able to form tissue bound residues. Mild acid treatment could release a compound - AOZ - from the tissue bound residues. The application of an HPLC method for the detection of AOZ showed that it was both stable and persistent in the tissues of treated animals. This laboratory then developed a method for the detection of AOZ in pig kidneys^[49] and showed that AOZ could be detected in pig tissues for long periods after treatment with the drug had ceased^[50]. and that 16% of 200 pig kidneys in NI contained residues of AOZ^[50].

This confirmed that testing for furazolidone was ineffective. We had never found any furazolidone residues in NI pigs, but were able to show that it was widely used in the pig industry. When the ban was introduced, this laboratory was in a position to monitor compliance of the ban on the use of furazolidone - but was unable to control any of the other nitrofuran drugs. Subsequently, the EU-funded research project "FoodBRAND" (www.afsni.ac.uk/foodbrand) was able to develop a sensitive method for all four of the suggested metabolites of the nitrofuran drugs. These were: AOZ from furazolidone. AMOZ from furaltadone. AHD from nitrofurantoin and SEM from nitrofurazone. The method, which employs LC-MS-MS. meets the criteria set out by the EU for the unequivocal confirmation of the presence of these metabolites at concentrations well below 1.0 µg/kg^[51].

5. "Zero Tolerance", the MRPL and 3^rd countries

The development of methods for the detection of nitrofuran metabolites, at low concentrations has caused severe difficulties for a number of 3^rd countries, wishing to export food into Europe. Their controls on nitrofurans (where they existed) were exclusively based on the detection of nitrofuran parent drugs using HPLC-UV. The issue of Rapid Alerts and (in the cases of- for example - Thailand and Brazil) the adoption of protective measures^{[52], [53]} has had enormous effects on these countries in terms of trade restrictions, financial losses associated with rejected consignments and the cost of major capital investment in LC-MS/MS systems and associated staff. The EU has a policy of zero tolerance towards the use of nitrofurans in food-producing animals. According to the current legislation, any confirmed concentration of any of the metabolites is a non-compliance. EU Member States have condemned imported product when very low concentrations of the nitrofuran metabolites are detected and confirmed, providing that the competent authorities have had sufficient statistical confidence in the analytical method. Not unreasonably, 3^rd countries have made representations to the effect that a) EU MS take enforcement action at different thresholds; and b) that they, as exporting countries must de facto reach the same performance limits as the EU laboratory with the lowest (i.e. most sensitive) threshold.

Over all of the foregoing, the Minimum Required Performance Limit (MRPL), established under Commission Decision 2002/657/EC^[51], has cast a long shadow. The precise meaning of this term has caused much debate and confusion - inside and outside the EU since the promulgation of this Decision. Originally intended as a laboratory benchmarking tool - if a method can reliably and repeatedly confirm the presence of banned substances at concentrations equal to. or below, the MRPL - it would be deemed "fit for purpose". If it could not confirm at those concentrations -then further development work would be required to enable it to meet those levels. In effect it was there to put a "cap" on the worst performing methods to ensure that a minimum standard was applied across MS laboratories. The MRPL concept did not include any provision for a maximum standard. Nowhere is it stated that the lowest concentration of a banned substance requiring effective enforcement action is X.X µg/kg. If the banned substance can be confirmed with the required degree of statistical confidence (a-error of 1 %) - then the sample is non-compliant. Thus, the lowest concentration at which enforcement action could be taken - CCa - had the potential to vary markedly between laboratories - depending on the range of analytical equipment available. The upper limit (minimum standard) would be the MRPL. but the lower limit (maximum standard) could be considerably lower than the MRPL.

6. Constraints on finding a way forward

In the face of the near-global abuse of nitrofurans. it was inevitable that there would be conflicts of views and conflicts of interests - especially following the adoption of an MRPL for the nitrofuran metabolites of 1.0 µg/kg. Four statements (below) encapsulate the conflicts that have arisen.

1. Third counties, often with limited analytical capabilities, must be capable of trading their produce on the international market place. 2. Consumers and (more vocally) retail chains in Europe demand residue-free food. 3. Certain drugs were banned because it was impossible to assign an Acceptable Daily Intake of their residues. 4. Banned drugs will become legitimate!) and systematically abused though the adoption of "tolerance limits" that require no action at all when their presence below the "tolerance limit" is detected.

It is clear that there is a need for a more harmonised approach to international trade. It is also clear that banned substances were banned for a sound reason and that they should remain banned. What is required is that science and policy should work together to provide a framework to facilitate the control and gradual elimination of residues of banned substances from the food chain. However, this must be done in a manner that is compatible with the ability of 3rd countries to meet the analytical requirements of the EU and should also take into account the recent problems that have arisen with regard to the nitrofurazone metabolite, semicarbazide.

7. Semicarbazide

The FoodBRAND project developed screening and confirmatory tests to detect tissue-bound residues of the banned nitrofuran antibiotics. These methods have uncovered the widespread occurrence of nitrofuran metabolite residues in pigs, poultry and shrimps from a wide range of Asian, South American and European countries. Experiments in broilers, egg-layers and shrimp have shown that SEM is formed as a tissue bound residue following the administration of nitrofurazone. However, it has become apparent that SEM has been detected in a range of materials intended to coat chicken meat during the production of cooked chicken products. Most positive findings have been associated with the use of either carageenan or breadcrumbs (and other bread products). No obvious nitrofuran-related explanation could be found. However, a subsequent investigation revealed that azodicarbonamide, a flour treatment agent not permitted for use in the EU, was commonly used in the production of certain breaded chicken products produced in Thailand and elsewhere. It is now known that azodicarbonamide can break down during the bread making process to yield semicarbazide. This emphasised the importance of removing the coatings from chicken (and shrimp) products and of analysing the meat for the presence of bound residues of semicarbazide. Advice to do this was published on the Internet by the FoodBRAND team in May 2003 and was subsequently adopted by the Eli CRT for nitrofurans (AFSSA, France) later that year. Azodicarbonamide was also implicated in causing SEM in baby food preparations as a result of its use as a blowing agent in the manufacture of tamper-evident seals in food containers. Other workers have focussed on the role of bleaching as a cause of SEM in carageenan. a seaweed extract used widely in food production world-wide. Since SEM continues to be abused in some countries, there is again a need for science and policy to work together to provide a means of enforcing the ban on the use of nitrofurazone by the development of alternative, improved markers of abuse.

8. New analytical developments

If there is a global determination to remove nitrofuran residues from food, it would be inappropriate to ignore the latest analytical developments. It is now becoming apparent that recent developments can help further to reduce the prevalence of nitrofuran residues and, at the same time, enable third countries with a more limited analytical capacity to put in place a more realistic monitoring strategy.

Firstly, it is now apparent that, in a manner similar to the ß-agonists - the nitrofuran metabolites accumulate in retinal tissue to very high concentrations. In pigs (and depending on the nitrofuran concerned) retina concentrations 6 weeks after withdrawal of medication range from 0.1 (AHD - from nitrofurantoin) to 13.8 mg/kg (AMOZ -from furaliadone). These concentrations, which are several orders of magnitude higher than the corresponding concentrations in more conventional matrices (liver, kidney, muscle, etc) are sufficiently high that laboratories equipped only with HPLC-UV should be able to detect nitrofuran use in food animals for prolonged periods after cessation of treatment. If these findings can be repeated in other species (a trial involving nitrofurazone in poultry is already under way) - this represents a major analytical advance. It should be noted that the use of retina as a matrix for testing is already adopted in many countries for the detection of ß-agonist abuse, as these compounds can persist in cattle retina for at least one year after treatment. Retina testing has formed the cornerstone of Northern Ireland's programme to eliminate ß-agonist abuse in beef cattle, which was endemic in this region in the early 1990s.

Secondly, it is now apparent that nitrofurans accumulate in eggs in the form of the parent drug. It is a quirk of the analytical procedure, developed as part of FoodBRAND, that any parent drug present in a sample is measured as its corresponding "metabolite". If an egg containing 100 ppb of furazolidone (parent drug) is analysed for the presence of total nitrofuran metabolites, then the standard acid hydrolysis and o-nitrobenzaldehyde derivatisation procedure will convert the furazolidone to AOZ. This means that it may be possible a) to detect nitrofurans in eggs as their parent drugs using simpler procedures and b) determine whether or not a finding of semicarbazide in an egg sample results from a processing problem or from nitrofurazone abuse. Further studies exploring these developments are currently under way.

9. Conclusions.

Controlling and eliminating the abuse of nitrofurans in food producing animals is a complex and (often) costly procedure. However, by utilising state-of-the-art information and appropriate detection technologies, it may be possible to further reduce and ultimately eliminate the use of these compounds globally. This will be to the advantage of the consumer and the producer and should facilitate the continuing international trade in food.