Previous Page Table of Contents Next Page


Working papers

Risk Assessments for substances without ADI/MRL - an overview

Dieter Arnold, Berlin, Germany

1. Introduction

Hundreds of substances are currently used as active principles in health products developed for the treatment of food producing animals. However, a recent comprehensive scientific evaluation - employing contemporary risk assessment principles - of the pharmacologically active substances used in veterinary medicinal products in the European Union, has shown that it is not necessary for the protection of human health to establish maximum residue limits (MRLs) for all of these substances - if they are used as approved by the competent authorities[4]. One can estimate that the number of pharmacologically active substances possibly requiring establishment of international Maximum Residue Limits for Veterinary Drugs (MRLVDs in the terminology of the Codex Alimentarius) may range below 200. Foods containing residues of any of these substances could potentially be a hazard to the health of consumers and could create problems in international trade. However, many of these substances have never been subjected to an international evaluation.

Since the introduction of contemporary methods of risk assessment, the FAO/WHO Expert Committee on Food Additives (JECFA[5]) has evaluated approximately 75 substances (counting groups of substances having the same Acceptable Daily Intake (ADI) and MRLVDs as one substance). For some substances (for a list see table 1) it was not possible for JECFA to allocate an ADI and to recommend MRLVDs. However, for the majority of substances evaluated so far, JECFA was able to propose MRLVDs to the FAO/WHO Codex Alimentarius Committee on Residues of Veterinary Drugs in Foods (CCRVDF) a subsidiary body to the FAO/WHO Codex Alimentarius Commission (CAC).

The establishment of maximum (residue) limits represents one of several standard options for risk managers to limit the presence of undesired substances in foods. MRLVDs are Codex standards. Only at the level of the CAC can Codex standards be adopted. CCRVDF plays a key role in the eight step elaboration procedure for Codex standards[6], however, it gave JECFA the mandate to propose draft MRLVDs on the basis of a risk assessment. The first discussion of the proposal of JECFA ("proposed draft MRLVDs") takes place in the CCRVDF at step three of the elaboration procedure.

A complete up to date overview of the official MRLVDs adopted by the CAC cannot be easily obtained without consulting the reports of all recent sessions since the Codex data base on MRLVDs for residues of veterinary drugs has been updated on 2 September 1999 for the last time and does also not provide full information on the marker residues on which the MRLs are based. Furthermore, substances without MRLVD drop out of the Codex system since the CCRVDF has not yet considered risk management options for such substances which were on the past agendas of JECFA but for which neither an ADI had been allocated nor MRLVDs had been recommended. Therefore, the databases of FAO and WHO on the output of JECFA (e.g., reports, toxicological and residue monographs, summary and history of the evaluations) remain an important source of reliable complementary information[7]. These data bases are regularly updated shortly after the output of a JECFA meeting has been published.

2. Significance of ADI and MRLVD

It is important to know the Codex definition of MRLVD in order to better understand JECFA's risk assessment approach. The definition can be found in the Procedural Manual of the CAC[8]:

"Codex maximum limit for residues of veterinary drugs[9],[10] (MRLVD) is the maximum concentration of residue resulting from the use of a veterinary drug (expressed in mg/kg or µg/kg on a fresh weight basis) that is recommended by the Codex Alimentarius Commission to be legally permitted or recognized as acceptable in or on a food.

It is based on the type and amount of residue considered to be without any toxicological hazard for human health as expressed by the Acceptable Daily Intake (ADI), or on the basis of a temporary ADI that utilizes an additional safety factor. It also takes into account other relevant public health risks as well as food technological aspects.

When establishing an MRL, consideration is also given to residues that occur in food of plant origin and/or the environment. Furthermore, the MRL may be reduced to be consistent with good practices in the use of veterinary drugs[11] and to the extent that practical analytical methods are available."

It should be noted that MRLVDs have to be based on an ADI according to this definition. The ADI is an estimate by JECFA of the upper limit of the amount of a veterinary drug, expressed on a body weight basis, that can be ingested daily over lifetime without appreciable health risk. Thus, the MRLVD is a specific "health based" standard and, therefore, as a rule, where there is no possibility to allocate an ADI there is usually no way for JECFA to recommend MRLVDs. Legislation of the European Union uses a nearly identical definition of MRL and the consequences for the recommendations of the Committee for Veterinary Medicinal Products (CVMP) of the European Medicines Agency (EMEA) - if an ADI cannot be allocated - are typically similar to those described for JECFA.

Summarising the basic characteristics of MRLVDs one could state:

3. Risk assessment of substances without JECFA ADI and Codex MRLVD

Risk assessment - as described and defined in the Procedural Manual of the CAC[13] is: "A scientifically based process consisting of the following steps:

(i) hazard identification,

(ii) hazard characterization,

(iii) exposure assessment, and

(iv) risk characterization." where risk is defined as: A function of the probability of an adverse health effect and the severity of that effect, consequential to a hazard(s) in food.

3.1 Hazard Identification[14]

It is not possible to completely separate the hazard identification and hazard characterisation steps. However, the usefulness of a JECFA evaluation of the active principle of a veterinary drug depends critically on the quality of the identification of the hazard. Therefore, the information required is substantial and ranges from valid, well designed animal-based toxicity studies conducted in accordance with internationally recognized protocols (OECD) under conditions of Good Laboratory Practice (GLP) to validated in vitro techniques and even established structural alerts or structure activity relationships. "Well designed" means primarily, adequate selection of species, strain/genetic make up, sex, stage of life, number of animals per dose group and statistical design, adequate selection and spacing of doses possibly allowing the determination of a No Observed Effect Level (NOEL), suitable route of administration (per oral route, preferably via the diet).

For hazard identification of veterinary drugs requiring international evaluation typically a defined set of animal tests covering ideally the entire range of possible adverse effects are routinely required irrespective of the expected exposure of consumers. Information on dose dependent toxicokinetic properties, including rate and extent of absorption, distribution, pre-systemic and systemic metabolism and excretion in test animals, in the target species of animal (and - where available - in humans), is always indispensable for hazard identification as well as for hazard characterisation and may help to design and interpret other studies. If metabolites are formed in the target species of animal which were not observed in the laboratory animals separate testing of those metabolites may be necessary. Other special studies may be required on a case by case basis.

JECFA has developed a special approach for antimicrobial substances to identify and characterise antimicrobial hazards which could have adverse effects on the microbial flora of the gut and to determine an antimicrobial ADI[15], typically on the basis of specifically designed in vitro studies. An equivalent approach is used in the European Union. The two ADI figures - toxicological and antimicrobial ADI - are compared and a decision is made which one is more suitable to protect the health of consumers. From time to time observations in humans may be available if an active ingredient is also used in human pharmacotherapy.

Often overlooked is the fact that the results of any evaluations are strictly applicable only to a specified test substance used throughout the tests. Therefore, the availability of the results of specification of the test substance is important, in particular information on purity/impurities and on isomeric composition.

The data requirements of JECFA[16] are equivalent to those laid down in the legislation[17] and in guidance documents[18] of the competent institutions of the European Union.

Hazard identification is not a static science. Methods are constantly being re-assessed and improved and complemented by validated alternative methods. At the same time the data base on the biological potential of certain substances enlarges over the years. In particular the mandatory review of all active principles used in medicines for food producing animals in the European Union within less than one decade has encouraged sponsors to fill certain gaps of knowledge required for hazard identification. In other cases, however, substances which are still in use in other regions of the world have no longer been supported by sponsors or can no longer be used in food producing species in cases where it was not possible to establish MRLs. Scientific progress and other factors, e.g. changes in risk assessment policy suggest re-evaluation of all substances at suitable intervals.

Many substances came into use as veterinary drugs years before the development of modern techniques and standards for risk assessment. JECFA has adopted an approach to the evaluation of such "substances with a long history of use". It implies that for the assessment of such substances and on a case by case basis certain studies not meeting modern criteria could be used (or even be replaced) in the light of other equivalent information, presented for example, in the form of literature reviews conducted by sponsors and/or evaluation reports drafted by experts in the field, provided that the safety of these products can be assured to an extent equivalent to that achieved for newer products[19] supported by a contemporary base.

At the end of the hazard identification the effects which are considered adverse and the identity of the molecules which could cause these effects are known.

3.2 Hazard Characterisation[20]

In the risk assessment of veterinary drugs this step primarily focuses on the dose-response relationship for critical adverse effects, identification of the most sensitive species/strain for a given effect, and extrapolation from animals to humans. Furthermore any information elucidating the key processes involved in toxicity is important. Again toxicokinetic data may play an important role. They may, for example, provide information on the formation of critical metabolites and binding to cellular macromolecules and explanations for quantitative species differences and inter-individual variability.

JECFA's approach to use dose-response information in hazard characterisation bases on the traditional assumption[21] that non-cancer endpoints and non-genotoxic cancer endpoints display dose thresholds. If a substance causes only threshold responses in the exposed population of test animals the Committee determines for every critical effect the dose level at which there are no significant increases in frequency and severity of the effect between the exposed groups and the control groups. This dose level is called No Observed Effect Level (NOEL). The lowest NOEL or highest dose which does not cause any toxicity is typically, but not always taken as the starting point for deriving the ADI. The choice is also determined by the required "safety factors" (SF) by which the NOEL is divided in order to take account of the small size of the test group of animal, the inter-individual and inter-species variability, the nature of the critical effects and the quality of the data. In fact, the combination of NOEL and SF resulting in the lowest intake (and thereby the largest margin of safety for all critical effects observed in the exposed animals) is finally used to allocate the ADI. Evaluations carried out in the EU principally follow the same approach.

This approach appears to be simple and pragmatic, however, depending on factors such as, the slope of the dose response relationship, dose spacing, number of animals, etc., the NOEL estimate may lack accuracy and may not always represent a true sub-threshold value. The proportion of the human population protected (not including sensitive subpopulations) is not exactly known. The JECFA ADI is always presented as a range from zero to an upper limit encouraging efforts to keep human exposure as low as GPVD allows. The ADIs of the CVMP are usually presented as one figure. Since mathematical modelling and low-dose extrapolations of dose response relationship is not used by JECFA, valuable information is lost and no quantitative estimate of risk at any level of intake, particularly not at intake levels above the NOEL/ADI can be provided to risk managers.

For effects involving non-threshold mechanisms, the NOEL/SF approach cannot be used. Both, JECFA and the CVMP/EMEA have not yet developed an alternative approach (which would also need to be accepted by their respective risk managers [CCRVDF and Commission/Council of the EU, respectively.] to deal with substances exhibiting such effects. The above cited definition of MRLVD does currently not allow approaches other than the ADI approach. Therefore, substances exhibiting non-threshold effects drop out of the system irrespective of the magnitude of risk posed to consumers at very low levels of contamination of foods. Theoretically, if suitable incidence data of the relevant effects observed in the animal studies are available they could be used for low dose extrapolation and risk estimation and determination of a "virtually safe dose". However, no mathematical model which has been proposed in the past decades has found full international scientific acceptance in the field of veterinary drugs. The International Programme on Chemical Safety (IPCS) has announced an Expert Consultation on the Harmonization of the Application of Dose-Response Modelling in Chemical Risk Assessment to provide specific advice on this issue.[22]

At the end of the hazard characterisation step performed by JECFA, the upper limit of the range of acceptable daily intakes of the hazardous agent is known. It is assumed that intakes over lifetime at or below this limit would not pose an appreciable risk to the health of consumers.

3.3 Exposure Assessment[23]

In the evaluation of the active principles of veterinary drugs it suffices to carry out an assessment of dietary intakes and to compare the result with the Acceptable Daily Intake. The minimum amount of information required to perform an assessment of intake of residues is:

Of particular interest is also an estimate of the probability of consuming occasionally high amounts containing at the same time high concentrations of residues. The information required under the first point is very complex and includes:

This list is not exhaustive and other studies may, in some instances, be of assistance in the evaluation of residues.

Concerning the second point, CCRVDF has unsuccessfully attempted during its early sessions to get information from Member States about dietary intakes in order to provide this information to JECFA. Therefore it has been decided to use arbitrary intake figures ranging close to the upper limits of daily intakes by consumers of edible tissues and animal products. Currently the Committee uses the following intake figures: 300 g of muscle (muscle and skin in natural proportions in the case of fish), 100 g of liver, 50 g of kidney, 50 g of tissue fat, 100 g of egg, 1.5 L of milk, and 20 g of honey. In the case of pigs and of poultry and depending on the nature of the substance, fat may be replaced by the same amount of fat and skin in natural proportions.

JECFA has adopted a procedure by which a numerical value called "Theoretical Maximum Daily Intake" (TMDI) is calculated and compared with the ADI. It is assumed that a person with a body weight of 60kg consumes every day over lifetime 500 g of mammalian or poultry "meat" (muscle, liver, kidney, fat) or 300 g of fish as detailed above plus 1500 g milk plus 100 g egg plus 20g honey. Furthermore it is assumed that all the above tissues and products contain residues at a concentration equivalent to the MRLVD every day. For all these tissues and products the fictive consumption is multiplied with the corresponding tentative MRLVD and the resulting quantities of residues are summed up to yield the TMDI. Initially the TMDI was introduced as a first crude screening in a tiered approach because a more refined and more realistic estimate was unnecessary as long as the TMDI was blow the ADI. The TMDI greatly overestimates the true intakes of residues. However, over the years it became practice of the risk managers in CCRVDF as well as in the European Union not to accept proposed MRLVDs if the TMDI was markedly above the ADI.

MRLVDs are usually set at the lowest concentrations which are compatible with the ADI, the data on the depletion of the residues, the established GPVD including withdrawal times (as far as information is available) and the performance of proposed regulatory analytical methods. The MRLVDs cannot be lower than the concentrations of marker residue remaining at the end of the approved withdrawal period. Usually the upper limit of the 95% confidence interval over the 95th or 99th percentile of residue concentrations found in treated animals is used for comparison.

If a set of MRLVDs would result in a TMDI significantly above the ADI and no further reduction of the MRLVDs is practicable, no MRLVDs are proposed by JECFA.

3.4 Risk Characterisation[24]

Risk characterisation is the step in risk assessment in which the results of hazard characterisation and exposure assessment are integrated and transformed into advice to risk managers. The final output and the complexity of the risk characterisation step depends on the mandate given to risk assessors. The desired output in the case of veterinary drugs consists primarily of a set of proposed draft MRLVDs if a substance has successfully passed the evaluation. In this case this output is explained in a public report justifying the selection of the proposed MRLVDs on the basis of an accurate description and discussion of the results of the three steps, hazard identification, hazard characterisation and exposure assessment. If no ADI and/or MRLs could be recommended, this is also explained and if gaps in information were the reason then the data requirements to be fulfilled before a further review are defined.

Due to the described limitations of the established risk assessment procedures managers will not obtain:

This weakness of the system makes it occasionally difficult for managers to take proportionate measures if "illegal" residues are found.

4. Substances without ADI and MRLVD

A substance may not have MRLVDs for a variety of reasons some of which are briefly summarised below. The flow chart in figure 1 shows some examples of critical studies and decision points in the evaluation process of JECFA.

- The substance

- The hazard identification step revealed a potential for non-threshold effects. An ADI was not allocated. The procedure was discontinued as consequence of risk assessment policy.

- An ADI was allocated. However, all sets of tentative MRLVDs which were compatible with practical withdrawal periods and/or analytical methods resulted in an excess over the ADI of the TMDI. The procedure was discontinued as a consequence of risk assessment policy.

Figure 1 Examples of some decision points in the establishment of proposed draft MRLVDs

Table 1 lists substances which have been evaluated by JECFA and which currently have no ADI and no MRL. The outcome of JECFA evaluations is compared with the outcome of evaluations carried out in the European Union, since both sides developed their risk assessment approaches about a the same time. Table 1 starts with a number of substances (1-9) which JECFA evaluated long before the contemporary concepts evolved. The ADI concept had not yet been developed for antimicrobial substances. Therefore substances 2.-9. have no JECFA ADI, but (partly based on new data) EU ADIs and MRLs as far as they have been evaluated. For Benzyl penicillin both sides established MRLs without allocating a "classical ADI" (based on NOEL/SF approach). In a few other cases where CVMP allocated ADIs, the two Committees did not always work on the same data base or CVMP was more flexible in bridging gaps in the information.

In all cases were JECFA expressed safety concerns, the outcome of the CVMP evaluation was also negative and resulted in an inclusion into Annex IV of Council Regulation 2377/90 (prohibition of the use in food producing animals). The EU is not the only region where the use of such substances has been prohibited. The same has happened - sometimes years earlier - in other regions of the world. The European Union was taken as example because the JECFA system and the EU system evolved in the same decade.

Table 1 Substances without JECFA ADI

Substance

Outcome of JECFA evaluations

Outcome of EU evaluations

Year of last JECFA evaluation

Comment

ADI of the CVMP/EMEA
[µg/kg bw/day]

Comment

1. Diethylstilboestrol

1960

Substance is carcinogenic


prohibiteda)

2. Bacitracin

1968

JECFA recommended "Acceptable Residue Levels"

3.9

Annex Ib),c)

3. Erythromycin

1968


5

Annex Ib)

4. Novobiocin

1968


1.25

Annex Ib)

5. Kanamycin

1968


8

Annex Ib)

6. Leucomycin

1968



f)

7. Nystatin

1968

JECFA recommended "Acceptable Residue Levels"


f)

8. Oleandomycin

1968



f)

9. Polymixin B

1968



f)

10. Benzyl-penicillin

1990

JECFA recommended MRLVDsd)


Annex Ib)

11. Carbadox

2003



c)

12. Cefuroxim Sodium

2004



f)

13. Chloramphenicol

2004



Annex IVb)

14. Chlorpromazine

1991



Annex IVb)

15. Dimetridazole

1989



Annex IVb),c)

16. Furazolidone

1992



Annex IVb)

17. Ipronidazole

1989



c)

18. Metronidazole

1989



Annex IVb)

19. Nitrofurazone

1992



Annex IVb)

20. Olaquindox

1994

Temporary acceptance


c)

21. Oxolinic Acid

1994


2.5

Annex I 2)

22. Permethrin (cis-/trans-isomers 80:20)

2000


10

Annex I 2)

23. Propionylpromazine

1991



f)

24. Ronidazole

1994



Annex IVb),c)

25. Sulfathiazole

1989



Annex Ib)

26. Tylosin

1991


6

Annex Ib),c)

27. Xylazine

1996



Annex IIb),e)

a) prohibited by specific legislation since 1981; currently in force: Directive 96/22 (EC) as amended by Directive 2003/74 (EC)

b) Refers to the Annexes of Council Regulation (EEC) No 2377/90 of 26 June 1990 laying down a Community procedure for the establishment of maximum residue limits of veterinary medicinal products in foodstuffs of animal origin as amended by Commission Regulation (EEC) No 762/92 of 27 March 1992 and Council Regulations (EC) 434/97 of 3 March 1997 and No 1308/1999 of 15 June 1999. Annex I lists the substances with MRL; Annex II lists the substances for which "...it appears that it is not necessary for the protection of public health to establish a maximum residue limit..."; Annex IV lists the substances for which "...it appears that a maximum residue limit cannot be established in respect of a pharmacologically active substance used in veterinary medicinal products because residues of the substances concerned, at whatever limit, in foodstuffs of animal origin constitute a hazard to the health of the consumer..."

c) Uses as veterinary medicinal products and as feed additives, respectively, are regulated by different legislation. Earlier authorization as feed additive have been withdrawn in recent years.

d) JECFA recommended an upper limit of daily dietary intake which was not derived from an NOAEL-Safety factor approach.

e) New data on the metabolism were available to CVMP which had not yet been available to JECFA.

f) Status not known to the author; no summary report published on the web site of the EMEA.

5. Summary


[4] See Annex II to Council Regulation (EEC) No 2377/90 of 26 June 1990 laying down a Community procedure for the establishment of maximum residue limits of veterinary medicinal products in foodstuffs of animal origin. An unofficial consolidated list (last update July 2003) can be found on the web site of the European Commission under: http://pharmacos.eudra.org/F2/mrl/conspdf/MRL%20consol%202003-07-22%20EN.pdf
[5] In this paper statements about "JECFA" refer only to the ad hoc Committees evaluating veterinary drugs and are not necessarily always true for the ad hoc Committees evaluating food additives and contaminants.
[6] See page 19 of the English version of the Procedural Manual of the CAC (13th edition) which can be found under: http://www.codexalimentarius.net/web/proceduralmanual.jsp
[7] Toxicological Monographs and Summary of the evaluations see:
http://www.inchem.org/pages/jecfa.html
Residue Monographs see: http://www.fao.org/es/esn/jecfa/jecfavetdrugen.jsp
Reports see: http://www.who.int/ipcs/publications/jecfa/en/
[8] http://www.codexalimentarius.net/web/proceduralmanual.jsp
[9] Veterinary drug means any substance applied or administered to any food producing animal, such as meat or milk producing animals, poultry, fish or bees, whether used for therapeutic, prophylactic or diagnostic purposes or for modification of physiological functions or behaviour.
[10] Residues of veterinary drugs include the parent compounds and/or their metabolites in any edible portion of the animal product, and include residues of associated impurities of the veterinary drug concerned.
[11] Good Practice in the Use of Veterinary Drugs (GPVD) is the official recommended or authorized usage including withdrawal periods, approved by national authorities, of veterinary drugs under practical conditions.
[12] see: http://www.wto.org/english/docse/legale/15sps01e.htm
[13] See page 44 and pages 51-52 of the Procedural Manual (13th edition, English version)
[14] Codex definition: "The identification of biological, chemical, and physical agents capable of causing adverse health effects and which may be present in a particular food or group of foods (Hazard: A biological, chemical or physical agent in, or condition of, food with the potential to cause an adverse health effect)."
[15] Spiramycin was the first substance where such an approach was used (see 38th report, WHO Technical report Series 815, available on line under: http://www.who.int/ipcs/publications/jecfa/en/). The approach has since been refined and improved several times and is used for the evaluation of all substances having antimicrobial activity.
[16] See for example 42nd report of the Committee, WHO Technical report series 851, available on line under http://www.who.int/ipcs/publications/jecfa/en/
[17] See Annex V of Council Regulation (EEC) No 2377/90 of 26 June 1990 laying down a Community procedure for the establishment of maximum residue limits of veterinary medicinal products in foodstuffs of animal origin. Available on line under:
http://www.europa.eu.int/servlet/portail/RenderServlet?search=DocNumber&lg=en&nb_docs=
25&domain=Legislation&coll=&in_force= NO&an_doc=1990&nu_doc=2377&type_doc=Regulation

[18] The rules governing medicinal products in the European Union. Volume 8. Notice to applicants and Note for guidance. Veterinary medicinal products. Establishment of maximum residue limits (MRLs) for residues of veterinary medicinal products in foodstuffs of animal origin. Available on line under: http://pharmacos.eudra.org/F2/eudralex/vol-8/pdf/Vol8rev1Final24June2003.pdf
[19] The details of this approach are outlined in the 40th report of the Committee, WHO Technical Report Series 832, available on line under: http://www.who.int/ipcs/publications/jecfa/en/
[20] Codex definition: The qualitative and/or quantitative evaluation of the nature of the adverse health effects associated with biological, chemical and physical agents which may be present in food. For chemical agents, a dose-response assessment should be performed. For biological or physical agents, a dose-response assessment should be performed if the data are obtainable (Dose-Response Assessment: The determination of the relationship between the magnitude of exposure [dose] to a chemical, biological or physical agent and the severity and/or frequency of associated adverse health effects [response]).
[21] The existence of a threshold can not be proven from the data of a dose-response relationship.
[22] 13 - 17 September 2004 Geneva, Switzerland IPCS Expert Consultation on the Harmonization of the Application of Dose-Response Modelling in Chemical Risk Assessment
[23] Codex Definition: The qualitative and/or quantitative evaluation of the likely intake of biological, chemical, and physical agents via food as well as exposures from other sources if relevant.
[24] Codex definition: The qualitative and/or quantitative estimation, including attendant uncertainties, of the probability of occurrence and severity of known or potential adverse health effects in a given population based on hazard identification, hazard characterization and exposure assessment (Risk Estimate: The quantitative estimation of risk resulting from risk characterization).

Previous Page Top of Page Next Page