Toxicology Letters, 64f65 (19921783-787 0 1992 Elsevier Science Publishers B.V., All rights reserved 03784274/92/$5.00
783
Safety/risk assessment of pesticides: principles, procedures and examples Frank C. Lu” and Michael L. Doursonb ‘Consulting Toxicologist, Miami, FL (Formerly, Chief; FoodAdditives Unit, World Health Organization) and bSystemic Toxicants Assessment Branch, Environmental Criteria and Assessment, U.S. Environmental Protection Agency, Cincinnati, OH (USA)
Key words: ADI; DDT, Pesticides; RfD; Risk assessment
SUMMARY The principles and procedures for the assessment of the safety/risk of chemical used by the relevant WHO and EPA expert groups are outlined. The assessment in terms of acceptable daily intakes (ADIS) and reference doses (RfDs) of 25 pesticides is listed. The pesticides assessed are acephate, alachlor, amitrole, azinphos-methyl, benomyl, biphenthrin, bromophos, chlordane, chlorthalonil, cyhalothrin, DDT, EPTC, ethion, folpet, fosetyl-al, glyphosate, isofenphos, methomyl, methyl mercury, paraquat, phosphamidon, systhane, terbutyn, tribultyltin oxide, and vinclozin. In addition, their critical effects, the no-observed-effect levels and the size of the safety/uncertainty factors used are also listed to illustrate the diversity of the toxic effects and the resulting assessments. Furthermore, the enormous amount of data reviewed and the complex scientific judgement involved are also indicated. Considering the various uncertainties existing, the ADIs and RfDs do not differ appreciably in most instances. However, marked differences exist between the ADIs and RfDs of DDT and chlordane. It is suggested that re-evaluation be done on these, and perhaps other, chemicals.
INTRODUCTION
For the protection of the health of consumers, the World Health Organization (WHO) has been engaged in conjunction with the Food and Agriculture Organization of the United Nations (FAO) in the assessment of the potential health hazards associated with pesticide residues in food. Over the past 30 years, some 200 pesticides have been evaluated. Many of them re-evaluated repeatedly in light of new toxicological and other relevant data. The views in this paper are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency. The U.S. Government has the right to retain a nonexclusive royalty-free license in and to any copyright covering this article. Correspondence to: F.C. Lu, 7452 SW 143 Ave., Miami, FL 33183, USA.
784
The first Joint FAO/WHO Meeting on Pesticide Residues was held in 1961 111. The Reports and Monographs which contain summaries of relevant studies reviewed, comments on their relevance, and the evaluation arising from subsequent annual meetings are listed in a recent WHO document for ready reference [Zl. The U.S. Environmental Protection Agency (EPA) has also been actively engaged in the assessment of pesticides. These assessments as well as the summaries of the principal and supporting studies are incorporated in EPA’s Integrated Risk Information System (IRIS) [31. PRINCIPLES
All pesticides, because of their intended use, are toxic to some form of life. At certain levels of exposure, they are also toxic to humans. To accomplish the purposes of permitting their agricultural usage and to minimize their risks to consumers, their safety/risk must be individually assessed, in view of their marked difference in toxicity. The assessment involves (1) identification of the hazard, (2) determination of the critical effect and the no observed adverse effect level (NOAEL) , and (3) estimation of acceptable daily intake (ADI) or reference dose (RfD). The AD1 and RfD form the basis of assessing the safety/risk of the intake of a pesticide when it is compared to these assessed figures. In other words, when the intake is at or below the ADI/RfD, it may be considered “safe” or involving only an acceptable risk. PROCEDURES
As noted above, a pesticide generally exhibits a variety of toxic effects under diverse experimental conditions. To assure a comprehensive assessTABLE I CLASSIFICATION OF BIOLOGICAL DATA Biochemical aspects: absorption, distribution, elimination (and storage); biotransformation; effects on biochemical parameters. Special studies: pharmacology, potentiation, neurotoxicity, reproductive function, teratogenicity, carcinogenicity (if not part of long-term studies), mutagenicity, toxicity of reaction products and impurities, etc. Acute toxicity studies: LDM, signs of acute toxicity. Short-term toxicity studies. Long-term toxicity studies: studies covering the entire life span or over 50% of it. Observations in man: biochemical studies; clinical and epidemiological studies.
ment, it is essential to review all relevant data. These data are often categorized under several headings as shown in Table I. From these data, it is often possible to identify, from an array of effects, the critical effect which is the most sensitive indicator of toxicity. A NOAEL can then be assessed based on the dose-response relationship relative to this effect. The critical effects are usually identified from long-term toxicity studies. However, other studies, especially those on reproductive functions, teratogenicity, carcinogenicity, have also provided the principal data. Some of the critical effects and the representative pesticides are listed in Table II. The NOAEL denotes the level of exposure that entails no observed adverse effect under the experimental conditions. However, even with a NOAEL derived from studies carried out in a limited number of humans, a “safety factor” (SF, usually 10) is applied to arrive at the ADI. This is because of the differences in susceptibility among humans. EPA uses the term “uncertainty factor” (UF) indicating the uncertainties involved in assessing an RfD. With a great majority of the pesticides evaluated, the assessment is based on data derived from animal studies. In such cases, a larger SF/UF (usually 100) is applied to the NOAEL. This larger factor is considered necessary, because (in addition to the intraspecies variability among humans) humans are considered more susceptible than the experimental animals when compared on a per kg body weight basis. Furthermore, even larger factors are used when the critical effect is of a serious nature (e.g. amitrole: secondary tumorigenesis, tributyl tin oxide: immunotoxicity) or when the database is less than fully adequate. Once an ADI/RfD has been established for a pesticide, the safety/risk of consuming its residues in food can be assessed. This can be done by chemical analysis which is more accurate but more tedious. Several simple methods are available. The simplest involves summing the products of the per capita consumption of each of the food commodity and the permissible level of the residue in that commodity. This simple calculation, while representing gross estimation of the actual intake, is useful when it is smaller than the ADI/RfD, thereby assuring safety of the pesticide, and avoiding any further calculations. However, when it exceeds the ADI/RfD, refined calculations may be done taking into account the losses of residues on storage, processing and cooking of the food, and the actual use patterns of the pesticide, etc. EXAMPLES
A variety of critical effects and representative pesticides are listed in Table II. For these pesticides a SF/UF of 100 was applied to the NOAELs to arrive at their ADIs/RfDs.
786 TABLE II
THE VARIETY OF CRITICAL EFFECTS AND REPRESENTATIVE PESTICIDES 1. AChE inhibition 2. Anemia -
ethion, phosphamidon
alachlor, terbutyn
3. Cardiomyopathy -
s-ethyldipropyl thiocarbamate
4. Developmental toxicity 5. Kidney lesions 6. Lung lesions -
benomyl, glyphosate
chlorothalonil, methomyl paraquat
7. Nervous system effects 8. Organ weight changes 9. Testicular damage -
biphenthrin vinclozin
fosetyl, systhane
TABLE III LARGER SAFETY FACTOR/UNCERTAINTY
FACTOR
Critical Effects
NOAEL
SF/UF
ADI/RfD
Amitrole
Goitrogenicity
0.025
1000S
0.00003*
Chlordane
Regional liver hypertrophy 0,055
1000U
0.00005R
Folpet
Keratosis in G.I.
10
1000S
0.01*
Paraquat
Proliferative lung lesion
1.6
400s
0.004*
Tributyl tin oxide
Immunotoxicity
0.025
1000U
0.0003R
Chemical
S=SF,U=UF,A=ADI,R=RfD.
Where adequate human data are available, a smaller SF/UF is used. For example, the satisfactory human data allowed the application of a SF of 10 in estimating the ADIs of acephate and bromophos by WHO. Another example is methyl mercury on which extensive human data were gathered through extensive epidemiological studies conducted during several tragic poisoning episodes. With this chemical, EPA used an UF of 10 and WHO applied a SF of 7 to the LOAEL (low observed adverse effect level). The use of a smaller factor on a LOAEL was considered justifiable because this level was based on a single most sensitive individual of a large exposed population. Furthermore, fish, the major food containing this chemical, is a valuable source of animal protein in many parts of the world. SFs larger than 100 were used by WHO in estimating the ADIs of amitrole, folpet and paraquat because they induced, respectively, thyroid tumors through a secondary mechanism, keratosis in the G.I. tract, and
787
proliferative lung lesions. A larger UF was used by EPA in assessing chlordane because of a lack of adequate reproductive study and adequate chronic study in a second mammalian species and because of the generally inadequately sensitive endpoints in the existing studies especially in view of its accumulation in the body. A larger UF was also used in assessing tributyl tin oxide because of its adverse effect on the immune functions (Table III). Within the 27 pesticides reviewed in this paper, the ADIs and RfDs are either identical or fairly similar. However, three of them differ appreciably. With chlordane, WHO considered the database as adequate, hence used a SF of 100, resulting in an AD1 of 0.0005 mg/kg versus an RfD of 0.00006 n&kg. With folpet, EPA did not consider the keratosis in the G.I. tract as significant, and therefore used an UF of 100 (RfD = 0.1 mg/kg versus an AD1 of 0.01 mg/kg). A 40-fold difference existed between the ADI (0.02 mg/kg) and the RfD (0.0005 mg/kg) for DDT. In arriving at the ADI, WHO used a SF of 10 from the NOAEL of 0.25 mg/kg from extensive observations in humans, whereas EPA applied an UF of 100 to the NOAEL of 0.05 n-&kg observed in an animal study. Differences of this magnitude might warrant a further review of DDT, and perhaps other pesticides. REFERENCES 1
2 3
WHO (1962) Principles governing consumer safety in relation to pesticide residues, Report of a Joint FAO/WHO Meeting of Experts on Pesticide Residues. WHO Tech. Rep. Ser. 240. WHO (1991) Summary of Toxicological Evaluations Performed by the Joint FAO/WHO Meeting on Pesticide Residues (JMPR). WHO/PCS/ 91.4, World Health Organization. EPA (1992) Integrated Risk Information System (IRIS). Online. U.S. Environmental Protection Agency, Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH.
783
Safety/risk assessment of pesticides: principles, procedures and examples Frank C. Lu” and Michael L. Doursonb ‘Consulting Toxicologist, Miami, FL (Formerly, Chief; FoodAdditives Unit, World Health Organization) and bSystemic Toxicants Assessment Branch, Environmental Criteria and Assessment, U.S. Environmental Protection Agency, Cincinnati, OH (USA)
Key words: ADI; DDT, Pesticides; RfD; Risk assessment
SUMMARY The principles and procedures for the assessment of the safety/risk of chemical used by the relevant WHO and EPA expert groups are outlined. The assessment in terms of acceptable daily intakes (ADIS) and reference doses (RfDs) of 25 pesticides is listed. The pesticides assessed are acephate, alachlor, amitrole, azinphos-methyl, benomyl, biphenthrin, bromophos, chlordane, chlorthalonil, cyhalothrin, DDT, EPTC, ethion, folpet, fosetyl-al, glyphosate, isofenphos, methomyl, methyl mercury, paraquat, phosphamidon, systhane, terbutyn, tribultyltin oxide, and vinclozin. In addition, their critical effects, the no-observed-effect levels and the size of the safety/uncertainty factors used are also listed to illustrate the diversity of the toxic effects and the resulting assessments. Furthermore, the enormous amount of data reviewed and the complex scientific judgement involved are also indicated. Considering the various uncertainties existing, the ADIs and RfDs do not differ appreciably in most instances. However, marked differences exist between the ADIs and RfDs of DDT and chlordane. It is suggested that re-evaluation be done on these, and perhaps other, chemicals.
INTRODUCTION
For the protection of the health of consumers, the World Health Organization (WHO) has been engaged in conjunction with the Food and Agriculture Organization of the United Nations (FAO) in the assessment of the potential health hazards associated with pesticide residues in food. Over the past 30 years, some 200 pesticides have been evaluated. Many of them re-evaluated repeatedly in light of new toxicological and other relevant data. The views in this paper are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency. The U.S. Government has the right to retain a nonexclusive royalty-free license in and to any copyright covering this article. Correspondence to: F.C. Lu, 7452 SW 143 Ave., Miami, FL 33183, USA.
784
The first Joint FAO/WHO Meeting on Pesticide Residues was held in 1961 111. The Reports and Monographs which contain summaries of relevant studies reviewed, comments on their relevance, and the evaluation arising from subsequent annual meetings are listed in a recent WHO document for ready reference [Zl. The U.S. Environmental Protection Agency (EPA) has also been actively engaged in the assessment of pesticides. These assessments as well as the summaries of the principal and supporting studies are incorporated in EPA’s Integrated Risk Information System (IRIS) [31. PRINCIPLES
All pesticides, because of their intended use, are toxic to some form of life. At certain levels of exposure, they are also toxic to humans. To accomplish the purposes of permitting their agricultural usage and to minimize their risks to consumers, their safety/risk must be individually assessed, in view of their marked difference in toxicity. The assessment involves (1) identification of the hazard, (2) determination of the critical effect and the no observed adverse effect level (NOAEL) , and (3) estimation of acceptable daily intake (ADI) or reference dose (RfD). The AD1 and RfD form the basis of assessing the safety/risk of the intake of a pesticide when it is compared to these assessed figures. In other words, when the intake is at or below the ADI/RfD, it may be considered “safe” or involving only an acceptable risk. PROCEDURES
As noted above, a pesticide generally exhibits a variety of toxic effects under diverse experimental conditions. To assure a comprehensive assessTABLE I CLASSIFICATION OF BIOLOGICAL DATA Biochemical aspects: absorption, distribution, elimination (and storage); biotransformation; effects on biochemical parameters. Special studies: pharmacology, potentiation, neurotoxicity, reproductive function, teratogenicity, carcinogenicity (if not part of long-term studies), mutagenicity, toxicity of reaction products and impurities, etc. Acute toxicity studies: LDM, signs of acute toxicity. Short-term toxicity studies. Long-term toxicity studies: studies covering the entire life span or over 50% of it. Observations in man: biochemical studies; clinical and epidemiological studies.
ment, it is essential to review all relevant data. These data are often categorized under several headings as shown in Table I. From these data, it is often possible to identify, from an array of effects, the critical effect which is the most sensitive indicator of toxicity. A NOAEL can then be assessed based on the dose-response relationship relative to this effect. The critical effects are usually identified from long-term toxicity studies. However, other studies, especially those on reproductive functions, teratogenicity, carcinogenicity, have also provided the principal data. Some of the critical effects and the representative pesticides are listed in Table II. The NOAEL denotes the level of exposure that entails no observed adverse effect under the experimental conditions. However, even with a NOAEL derived from studies carried out in a limited number of humans, a “safety factor” (SF, usually 10) is applied to arrive at the ADI. This is because of the differences in susceptibility among humans. EPA uses the term “uncertainty factor” (UF) indicating the uncertainties involved in assessing an RfD. With a great majority of the pesticides evaluated, the assessment is based on data derived from animal studies. In such cases, a larger SF/UF (usually 100) is applied to the NOAEL. This larger factor is considered necessary, because (in addition to the intraspecies variability among humans) humans are considered more susceptible than the experimental animals when compared on a per kg body weight basis. Furthermore, even larger factors are used when the critical effect is of a serious nature (e.g. amitrole: secondary tumorigenesis, tributyl tin oxide: immunotoxicity) or when the database is less than fully adequate. Once an ADI/RfD has been established for a pesticide, the safety/risk of consuming its residues in food can be assessed. This can be done by chemical analysis which is more accurate but more tedious. Several simple methods are available. The simplest involves summing the products of the per capita consumption of each of the food commodity and the permissible level of the residue in that commodity. This simple calculation, while representing gross estimation of the actual intake, is useful when it is smaller than the ADI/RfD, thereby assuring safety of the pesticide, and avoiding any further calculations. However, when it exceeds the ADI/RfD, refined calculations may be done taking into account the losses of residues on storage, processing and cooking of the food, and the actual use patterns of the pesticide, etc. EXAMPLES
A variety of critical effects and representative pesticides are listed in Table II. For these pesticides a SF/UF of 100 was applied to the NOAELs to arrive at their ADIs/RfDs.
786 TABLE II
THE VARIETY OF CRITICAL EFFECTS AND REPRESENTATIVE PESTICIDES 1. AChE inhibition 2. Anemia -
ethion, phosphamidon
alachlor, terbutyn
3. Cardiomyopathy -
s-ethyldipropyl thiocarbamate
4. Developmental toxicity 5. Kidney lesions 6. Lung lesions -
benomyl, glyphosate
chlorothalonil, methomyl paraquat
7. Nervous system effects 8. Organ weight changes 9. Testicular damage -
biphenthrin vinclozin
fosetyl, systhane
TABLE III LARGER SAFETY FACTOR/UNCERTAINTY
FACTOR
Critical Effects
NOAEL
SF/UF
ADI/RfD
Amitrole
Goitrogenicity
0.025
1000S
0.00003*
Chlordane
Regional liver hypertrophy 0,055
1000U
0.00005R
Folpet
Keratosis in G.I.
10
1000S
0.01*
Paraquat
Proliferative lung lesion
1.6
400s
0.004*
Tributyl tin oxide
Immunotoxicity
0.025
1000U
0.0003R
Chemical
S=SF,U=UF,A=ADI,R=RfD.
Where adequate human data are available, a smaller SF/UF is used. For example, the satisfactory human data allowed the application of a SF of 10 in estimating the ADIs of acephate and bromophos by WHO. Another example is methyl mercury on which extensive human data were gathered through extensive epidemiological studies conducted during several tragic poisoning episodes. With this chemical, EPA used an UF of 10 and WHO applied a SF of 7 to the LOAEL (low observed adverse effect level). The use of a smaller factor on a LOAEL was considered justifiable because this level was based on a single most sensitive individual of a large exposed population. Furthermore, fish, the major food containing this chemical, is a valuable source of animal protein in many parts of the world. SFs larger than 100 were used by WHO in estimating the ADIs of amitrole, folpet and paraquat because they induced, respectively, thyroid tumors through a secondary mechanism, keratosis in the G.I. tract, and
787
proliferative lung lesions. A larger UF was used by EPA in assessing chlordane because of a lack of adequate reproductive study and adequate chronic study in a second mammalian species and because of the generally inadequately sensitive endpoints in the existing studies especially in view of its accumulation in the body. A larger UF was also used in assessing tributyl tin oxide because of its adverse effect on the immune functions (Table III). Within the 27 pesticides reviewed in this paper, the ADIs and RfDs are either identical or fairly similar. However, three of them differ appreciably. With chlordane, WHO considered the database as adequate, hence used a SF of 100, resulting in an AD1 of 0.0005 mg/kg versus an RfD of 0.00006 n&kg. With folpet, EPA did not consider the keratosis in the G.I. tract as significant, and therefore used an UF of 100 (RfD = 0.1 mg/kg versus an AD1 of 0.01 mg/kg). A 40-fold difference existed between the ADI (0.02 mg/kg) and the RfD (0.0005 mg/kg) for DDT. In arriving at the ADI, WHO used a SF of 10 from the NOAEL of 0.25 mg/kg from extensive observations in humans, whereas EPA applied an UF of 100 to the NOAEL of 0.05 n-&kg observed in an animal study. Differences of this magnitude might warrant a further review of DDT, and perhaps other pesticides. REFERENCES 1
2 3
WHO (1962) Principles governing consumer safety in relation to pesticide residues, Report of a Joint FAO/WHO Meeting of Experts on Pesticide Residues. WHO Tech. Rep. Ser. 240. WHO (1991) Summary of Toxicological Evaluations Performed by the Joint FAO/WHO Meeting on Pesticide Residues (JMPR). WHO/PCS/ 91.4, World Health Organization. EPA (1992) Integrated Risk Information System (IRIS). Online. U.S. Environmental Protection Agency, Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH.
