Toxicological Evaluation and Limit Values for Nonylphenol, Nonylphenol Ethoxylates, Tricresyl, Phosphates and Benzoic AcidTable of contentsPrefacePrinciples for setting of limit values for chemical substancesNonylphenol and nonylphenol ethoxylates1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 8. TDI, health based limit values 9. C-value Tricresyl phosphates1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 7. Evaluation 8. TDI, health based limit values 9. C-value Benzoic acid1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 8. TDI, health based limit values PrefaceThis series of reports constitutes a part of the work related to the setting of health based limit values for chemical substances in air, soil and drinking water. In this report, the toxicological documentation for the setting of limit values for nonylphenol and nonylphenol ethoxylates, tricresylphosphates and benzoic acid are presented. For every substance, the following items are considered:,
The work has been carried out by the Institute of Food Safety and Toxicology, Danish Veterinary and Food Administration as a contract work for the Danish Environmental Protection Agency. The work has been followed by a Steering Committee who has contributed to the work with professional expertise, proposals and criticism: Linda Bagge, Chairman, Danish Environmental Protection Agency Poul Bo Larsen, Danish Environmental Protection Agency Erik Thomsen, Danish Environmental Protection Agency Hans Chr. Ellehauge, Danish Environmental Protection Agency Anders Carlsen, Medical Health Office for Viborg County Elle Laursen, National Board of Health Ole Ladefoged, Institute of Food Safety and Toxicology Elsa Nielsen, Institute of Food Safety and Toxicology Principles for setting of limit values for chemical substancesIn the following, the principles upon which the Danish Environmental Protection Agency bases the health based limit values, in the following referred to as limit values, for chemical substances are briefly outlined. For further and more specific information, the reader is referred to the references mentioned below. Purpose The purpose of setting limit values for chemical substances is to prevent health hazards in the human population caused by chemicals as pollut ants. The scientific method for setting of limit values comprises a hazard identification and hazard assessment which together with an exposure assessment constitute the risk assessment part in the proces of setting limit values. Selection of data Data concerning exposure and harmful effects of a chemical substance are collected from national and international criteria documents, monographs and original scientific literature. During the review of the data, the quality and reliability of the studies and research work are critically assessed. This is an important step since conflicting viewpoints regarding the hazards may be present. Unpublished data from industry or other sources are only seldom used, as such data have not been published in scientific journals and have not been subjected to critical review by other scientists. If adequate human data are available these are preferred as the basis for the assessment. For most substances however, human data are not adequate or available. In these cases, limit values are based upon data from experimental animal studies. When all the relevant data have been evaluated, the hazard considered most important - "the critical effect" - for setting the limit value, is identified. In this step it is assessed whether an effect should be considered as adverse and of relevance to humans. A substance may have different effects at different concentrations or doses. Generally, the effects are of more concern the lower the concentration or dose at which they occur, and the effect observed at the lowest concentration or dose often forms the basis for setting the limit value. Threshold chemicals, NOAEL or LOAEL The next step for assessment of a limit value is to identify the "no ob served adverse effect level" (NOAEL) which is the highest dose at which the critical effect was not observed or, in cases where a NOAEL cannot be identified, the "lowest observed adverse effect level" (LOAEL) which is the lowest dose at which the critical effect was observed. TDI /safety factors Having identified a NOAEL or a LOAEL, three "safety factors" (SF) are used to extrapolate from NOAEL or LOAEL to the tolerable daily intake, TDI (expressed in mg/kg b.w. per day) or the limit value for air, LVa;r, (expressed in mg/m3). The purpose of the safety factors is to take into account the fact that:
Thus in cases where a threshold value for the toxic effect is assumed. and a NOAEL or a LOAEL can be identified, the TDI or the LVa;r are obtained by the following calculation: Exposure routes In general, limit values for air are based upon data from inhalation studies and limit values for soil (LVso;,) and drinking water (LVdW) are based upon data from oral studies. However, if data for the relevant exposure route are not available, data from alternative exposure routes may be used as well, although it is realized that the degree of uncertainty may increase. This will then influence the value of the SF3. Analogy In cases where no data on harmful effects are available, an evaluation may be made upon the basis of data for related substances and a consequent increase in the value of the SF3. Non threshold chemicals For chemical substances where a threshold value for the toxic effect can not be assumed (i.e. genotoxic carcinogenic substances), the concept of lifetime risk is applied. Thus, for these potential carcinogenic substances, the TDI corresponding to a specific lifetime risk, is calculated upon the basis of animal studies by means of the "One Hit" model. A lifetime risk of 10-6 (life-time exposure to the dose that may lead to cancer for one in a million) is considered as tolerable. Exposure air, water, soil Having obtained the tolerable daily intake for a chemical substance, the limit values for drinking water and soil are calculated taking into account the daily exposure from the various media. The following exposure stan dard estimates for the various media are used in the calculation of limit values:
*For the soil exposure estimates, it has to be emphasized that these are based upon exposure scenarios which cover the most sensitive applications, e.g. domestic gardens, play grounds or kindergartens. To ensure that the total daily intake of a chemical substance from the various media does not exceed the tolerable daily intake, a certain percentage of the tolerable intake to the various media may be assigned (allocation). Limit values The limit value for soil and drinking water are obtained by the following calculations: *TDI or a percentage of the TDI (allocation) w: body weight for a child (10 kg) or an adult person (70 kg) C-value, quality criteria. Finally, the limit values are used as the basis for the setting of quality criteria for soil, drinking water, and air (C-values). In this step, other than health based viewpoints may be taken into account. This may include aesthetical factors such as odour (all media), discoloration (soil, drinking water), taste and microbial growth (drinking water). Furthermore, economic or political administrative factors may be taken into account. It has to be stressed that no ecotoxicological considerations are taken into account in the process of setting health based limit values. References Industrial Air Pollution Control Guidelines. Vejledning fra Miljøstyrelsen Nr. 9 1992. Ministry of the Environment, Denmark, Danish Environmental Protection Agency. Health Based Evaluations of Chemical Substances in Drinking Water. Vejledning fra Miljøstyrelsen Nr. 1 1992. Ministry of the Environment, Denmark, Danish Environmental Protection Agency. In Danish. Risk Evaluation of Contaminated Sites. Miljoprojekt Nr. 123 1990. Ministry of the Environment, Denmark, Danish Environmental Protection Agency. In Danish. January 1999 / final Evaluation of health hazards by exposure to Nonylphenol and nonylphenol ethoxylatesand estimation of limit values in ambient air, soil and drinking water Grete Østergaard The Institute of Food Safety and Toxicology Danish Veterinary and Food Administration Denmark 1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 8. TDI, health based limit values 9. C-value 1. General description1. General description 1.1 Identity Nonylphenol (NP) is the commercially most important member of the group of alkyl
phenols. The term "nonylphenol" represents a large number of isomeric compounds,
varying in the point of attachment of the nonyl group to the phenol molecule, and in the
degree of branching in the nonyl moiety. Commercially produced nonylphenols are
predominantly 4-nonylphenol (para-nonylphenol) with a varied and undefined degree of
branching in the alkyl group, while very little straight chain nonylphenol is present
(EU-RAR 1998). Molecular formula: NP: C15H24O
Structural formula:
1.2 Physical / chemical properties
1.3 Production and use Nonylphenol NP is manufactured by reacting mixed nonenes with phenol. The nonyl group, which may be branched or linear, may be linked to the ring either ortho, meta or para to the OH group. In the EU, 78,000 tonnes of nonylphenol were produced in 1994 (EU-RAR 1998). NP is used as a starting material in the synthesis of NPEs, and as a monomer in polymer production. Nonylphenol ethoxylate NPEs are manufactured by reacting NP with ethylene oxide. A polyethylene oxide chain of
any desired length can be built up by continued introduction of ethylene oxide into the
reaction mixture. Such a reaction yields NPEs with a mixture of ethylene chain lengths,
and the number of ethylene units used to describe the product is the average number
(Swisher 1970). During base catalysed ethoxylation of nonylphenol, ethylene oxide
preferentially reacts with the free nonylphenol and only when this has all reacted do
longer ethylene oxide chains form. Nonylphenol ethoxylates have much narrower homologue
distributions than alcohol ethoxylates (ICI datasheet for Synperonic NP surfactants). In the EU, 109,808 tonnes of nonylphenol ethoxylates were produced in 1994 (EU-RAR 1998). NPEs are non-ionic surfactants. They are used industrially, as ingredients in institutional cleaners and detergents, and in household cleaning and personal-care products. Industrially, NPEs are used for emulsion polymerisation and polymer stabilisation, textile processing, in agricultural chemicals, pulp and paper processing, metal and mineral processing, latex paints, wetting agents and emulsifiers, foaming agents, inks, adhesives, and pharmaceuticals (Anonymous 1997). 1.4 Environmental occurrence NP and NPE are not known to exist in nature. Air NP: NP has not been measured in the atmosphere (EU-RAR 1998). Water NP has been found in surface water, sea water, and ground water (data from Switzerland, United Kingdom, USA, and Croatia). A typical value was 1 m g/l, and the highest value measured was 180 m g/l (River Aire, UK) (EU-RAR 1998). NPEs have been found in European surface water in concentrations ranging from below the limit of detection to 59m g/l (Talmage 1994). Soil In sludge from waste water treatment plants, concentrations of NP in the order of gram per kg dry weight have been found. In soil treated with sludge, 4.7 mg/kg has been found (EU-RAR 1998). In river sediments, NPE concentrations of the order of mg per kg dry weight have been measured (Talmage 1994). Foodstuffs NP was detected in raw beef samples, concentration not stated (HSDB 1997). 1.5 Environmental fate NP is not readily biodegradable. Several mechanisms of microbial aromatic ring degradation have been reported, the most common being formation of catechol from phenol, followed by ring scission between or adjacent to the two hydroxyl groups (Talmage 1994). NPEs may degrade into nonylphenol. During degradation NPEs’ ethylene oxide units are cleaved off the ethylene oxide chain until only short-chain NPE remain, typically mono- and diethylene oxides. Oxidation of these oligomers creates the corresponding carboxylic acids. This leaves several degradation products: short-chain ethoxylates, their carboxylic acids, and nonylphenols (Anonymous 1997). The rate of biodegradation seems to decrease with increasing length of the ethylene oxide chain (Talmage 1994). Air NP released to the atmosphere will exist in the vapour phase and is thought to be degraded by reaction with photochemically produced hydroxyl radicals, with a calculated half-life of 0.3 days. No data have been found for NPE. Water Abiotic degradation of NP is negligible. Biodegradation does not readily take place. The half-life in surface water may be around 30 days (EU-RAR 1998). No data have been found for NPE. Soil NP in soil will have no mobility (HSDB 1997). Biodegradation of NP in soil is slow, and may occur in steps, each step characterised by a certain rate of degradation, as shown in field tests. The half-life in soil is probably around 30 days (EU-RAR 1998). No data have been found for NPE. Bioaccumulation NP bioconcentrates to a significant extent in aquatic species. Excretion and metabolism is rapid (EU-RAR 1998). No data have been found for NPE. 1.6 Human exposure The possible routes of human exposure to NP and NPEs are dermal contact and inhalation by workers involved in the manufacture and use, dermal and inhalation exposure of consumers from household pesticide products, dermal contact to cleaning products and cosmetics, mucous membrane contact to spermicides; inhalatory exposure via the environment through air, and oral exposure via the environment through drinking water and food. 2. Toxicokinetics2. Toxicokinetics 2.1 Absorption, distribution Inhalation No data have been found for either NP or NPE. However, on the basis that NP appears to be readily absorbed from the gastrointestinal tract and in view of its high partition coefficient (approximately 4.5) (EU-RAR 1998) it is assumed that significant absorption via the inhalation route will occur. Oral intake Knaak et al. (1966) administered 14C-labelled nonylphenol, 14C-labelled nonylphenol ethoxylate (9 ethylene oxide units), and 14C-labelled polyoxyethylene, orally or intraperitoneally to male rats (4 rats per group). Daily urine and faeces samples were collected and analysed for 14C over a 7-day period, while exhaled CO2 samples were collected and analysed for 14C over a 4-day period. For NP and NPE, a similar pattern of excretion was found. About 70% of the administered radioactivity was excreted via faeces, and 20% via the urine. Most of the radioactivity was excreted within 4 days after administration. No radioactivity was detected in exhaled CO2. The main part of the urinary NP and NPE metabolites were found to be acidic and were believed to be glucuronic acid conjugates of nonylphenol. The toxicokinetic behaviour of radiolabelled nonylphenol was investigated in two male volunteers, aged 29 and 58 years (Müller 1997 - quoted in EU-RAR 1998). Ring 14C-labelled nonylphenol was administered to one volunteer as a single oral dose of 5 mg (66 µg/kg bodyweight) and to the second volunteer as a single intravenous dose of 1 mg (14 µg/kg). Blood, urine and faeces were collected from the first volunteer at intervals for up to 56 hours after administration. Blood samples only were taken from the second volunteer, for up 24 hours. The biological samples were analysed for nonylphenol and nonylphenol conjugates (glucuronide and sulphate) by gas chromatography/mass spectrometry (it is not clear why radiolabelled nonylphenol was used). Recovery experiments using spiked blood, urine, faeces and adipose tissue samples confirmed the efficiency of the analytical extraction technique. Following oral administration, the concentration of nonylphenol and nonylphenol present as conjugates in the blood both peaked at about 1 hour; peak concentration of nonylphenol present as a conjugate was 86 ng/g blood, which was some 100-fold greater than that of unconjugated nonylphenol. For intravenous administration, the highest concentrations of nonylphenol, at 0.6 and 0.2 ng/g blood for unconjugated and conjugated compound, respectively, were seen at the first sampling point of 30 minutes; at all time points the concentrations of unconjugated and conjugated nonylphenol were of the same order of magnitude. For both the oral and intravenous routes, the time courses of blood concentration were indicative of an initial phase of distribution from the blood to a second compartment (presumably the lipid compartment) followed by a slower elimination phase. Comparison of the AUCs for the oral and intravenous routes suggested that oral bioavailability of unconjugated nonylphenol was about 20%. Analysis of the urine samples showed that about 10% of the oral dose was excreted in urine as unconjugated or conjugated nonylphenol, most of which was eliminated within eight hours. Only about 1.5% of the oral dose was excreted in the faeces during the 56-hour collection period. Müller (1997 - quoted in EU-RAR 1998) also measured the nonylphenol content of 25 samples of adipose tissue taken at autopsies of persons thought to have had no occupational exposure to alkyl phenols. The measured tissue concentrations were all within the range of background contamination found in the analytical "blank" samples. The author indicated that all reasonable precautions were taken to minimise contamination during analysis. Dermal contact No data have been found. On the basis that nonylphenol appears to be readily absorbed from the gastrointestinal tract, and in view of its high partition coefficient, and since dermal LD50-values have been determined for NP and NPE, it is assumed that significant absorption via the dermal route will occur. The limited dermal toxicity data that are available indicate that the oral and dermal LD50-values for nonylphenol are similar. 2.2 Elimination See above 2.3 Toxicological mechanisms There are no data on toxicological mechanisms for either NP or NPE with the exception of oestrogenic effects of NP. NP is believed to interact directly with the oestrogen receptor (Odum et al. 1997). 3. Human toxicity3. Human toxicity 3.1 Short term toxicity Inhalation No data have been found. Oral intake No data have been found. Dermal contact Cosmetic formulations containing NPEs with 4, 9, or 12 ethylene oxide units in varying concentrations have been tested for skin irritation in human volunteers (CIR 1983). The results show a range of effects from no to mild skin irritation. NPEs with 4, 9, 15, or 50 ethylene oxide units were non-sensitising in human test subjects (CIR 1983). Mucous membrane contact A 10% NPE spermicidal film caused mild local vaginal irritation in 2 out of 30 women (CIR 1983). 3.2 Long term toxicity No data have been found. 3.3 Reproductive / developmental effects APEs and NPEs (of ethoxylate chain length 9 and 11) are used as spermicides. The results of three large-scale epidemiological studies do not indicate a relationship between use of spermicides and delivery of malformed infants (IPCS 1998). 3.4 Genotoxic effects No data have been found. 3.5 Carcinogenic effects No data have been found. 4. Toxicity, animal data4. Toxicity, animal data 4.1 Short term toxicity Inhalation nonylphenol The LC50-value of NP is unknown. Four hours was the maximum survival time for rats inhaling a "concentrated vapour" of NP, the concentration was not stated (Smyth et al. 1962, 1969 - quoted from Talmage 1994). The sensory irritation potential of nonylphenol has been investigated (EU-RAR 1998). Atmospheres of saturated vapour concentration and one tenth saturated vapour concentration, nominally 3636 mg/m3 (400 ppm) and 267 mg/m3 (30 ppm), respectively, were tested. Groups of five female CD-1 mice were exposed, nose only, to each concentration and the respiration rate was monitored using pressure plethysmography. The duration of exposure to the nonylphenol vapour was not reported. The proportion of liquid particulate material in the test atmospheres was determined, and found to be approximately 1% of the nominal concentration, an amount considered unlikely to have a significant influence on the results. At 3636 mg/m3 a mean respiratory rate depression of about 25% was found during exposure. However, at 267 mg/m3 there were no changes in the respiratory rate. These results suggest that nonylphenol can cause mild sensory irritation to the respiratory tract at high exposure levels. nonylphenol ethoxylate The LC50-value for NPE is unknown, although several studies of acute inhalatory toxicity have been performed (Table 1). No further details about the concentrations of the test atmospheres were given in the reports.
Oral administration nonylphenol Three unpublished studies performed according to OECD test guideline 401 have yielded
oral LD50 values of 1200 to 2400 mg/kg for males, and 1600 to 1900 mg/kg for
females; presumably the test species was the rat (EU-RAR 1998). nonylphenol ethoxylate A number of oral LD50 values for NPE with various ethylene oxide chain lengths have been reported. These values are presented in Table 2.
Dermal contact For NP, a dermal LD50-value of 2031 mg/kg in rabbits has been reported
(EU-RAR 1998)
NP is corrosive on contact with skin and is a severe eye irritant. Exposure to the saturated vapour may lead to mild sensory irritation of the respiratory tract (EU-RAR 1998). 4.2 Long term toxicity Inhalation No data have been found. Oral administration NP 28-day study In a 28-day study (quoted in IUCLID 1996) performed according to OECD guideline 407 (in 1981), doses of 0, 25, 100 or 400 mg/kg/day of NP were administered to Sprague-Dawley rats in the diet. At the highest dose level, body weight, food consumption, and food utilisation was statistically significantly reduced in both sexes. Also at the highest dose level, for male animals only, relative kidney, liver and testes weights were statistically significantly increased (by 20%), blood urea and cholesterol levels were statistically significantly increased, and glucose was statistically significantly reduced. Histopathological examination revealed hyaline droplet accumulation in the renal proximal tubules, and a minor vacuolation in the periportal hepatocytes. Females did not show these effects. In the EU-RAR (1998), the NOAEL is considered to be 100 mg/kg/day. NP 90-day study In a 90-day study performed according to EPA guidelines and of GLP quality, Sprague-Dawley rats were administered NP in the diet at concentrations of 0, 200, 650, or 2000 ppm; corresponding to a calculated (in the report) intake of 0, 15, 50 or 150 mg/kg/day (Cunny et al 1997). At the highest dose level, body weight, food consumption, and food utilisation was statistically significantly reduced for both sexes. Haematology, serum chemistry, and ophthalmoscopy findings, oestrous cycle pattern, and spermatogenesis were not affected by treatment. In males, a statistically significant dose-related increase in absolute and relative kidney weight, without accompanying histopathological or clinical-chemical findings was found (actually, a decrease in the occurrence of hyaline droplets was found at the highest dose level). In females of the highest dose group, absolute ovary weight was slightly decreased, without accompanying histopathological changes. Relative ovary weight was not affected. Relative liver weight was increased by 10% in both sexes at the highest dose, and in males only at the next-highest dose level. The NOAEL was considered by the authors to be 50 mg/kg b.w./day. NP multigeneration study Further information on repeated dose toxicity can be derived from a good-quality
multigeneration study (NIEHS 1998 - quoted in EU-RAR 1998). This study is also described
in section 4.3. Groups of 30 male and 30 female Sprague-Dawley rats were exposed to
nonylphenol in the diet at concentrations of 0 (control) 200, 650 or 2000 ppm over three
generations. Calculated nonylphenol intakes were, respectively, about 0, 15, 50 and 160
mg/kg/day during non-reproductive phases. The F0 generation were exposed for 15 weeks, the
F1 and F2 generations from soon after birth to about 20 weeks of age and the F3 generation
from birth to about 8 weeks of age.
NPEs 90-day studies Smyth & Calandra (1969), reported on toxicity studies on alkyl phenol ethoxylates including NPEs. Ninety-day feeding studies in rats have been performed with NPEs with 4, 6, 9, 15, 20, 30, and 40 ethylene oxide units. Test materials were of commercial grade and were added to the diet. The various studies were performed at five different laboratories in the years 1959-65 using individual test protocols. Results are thus not directly comparable. Dose groups consisted of 10 male and 10 female rats, except for one study (Shelanski), where groups consisted of 15 rats of "mixed sexes". The results are presented in Table 5. In the five studies by Industrial Bio-Test laboratories (NPEs with 4, 6, 15, 20, or 30 ethylene oxide units; doses are shown in Table 5) haematology was studied in 5 rats of each sex from the highest dose level and the control group before treatment and after 11 weeks. All rats were studied for gross pathology. Livers, kidneys, and testes were weighed, and 33 tissues were examined histopathologically from 5 rats of each sex at the highest and control levels. In order to discriminate between poor palatability and toxic effect, 25-day paired-feeding studies were performed at dose levels for which weight gains were shown to be lower than those of the control groups. For NPEs with 4 or 6 ethylene oxide units, effects included growth retardation and increased absolute and relative liver weight. For NPEs with 15 or 20 ethylene oxide units only retarded growth was found. All effects on growth rate were judged to be due to poor palatability of the diets. With respect to the increased liver weights (for NPEs with 4 or 6 ethylene oxide units), no accompanying histopathological findings were found (however, only the highest dose level was examined histopathologically). The liver response was interpreted by the authors as an increase in parenchymatous tissue resulting from increased enzyme activity in relation to metabolism of the test substances. For NPE with 30 ethylene oxide units no effects at all were found. In the study of NPE with 9 ethylene oxide units by Mellon Institute gross pathology was studied on all rats at sacrifice, livers and kidneys were weighed, and 16 tissues were studied from 12 controls and from 8 rats from the highest dose group, while only 3 tissues were examined in 10 rats on the other two dose levels. Effects included growth retardation and increased relative liver weight at the two highest dose levels (250 or 1250 mg/kg/day). The liver weight increase was accompanied by cloudy swelling, intracellular lipoid, and reduced polysaccharide, while focal hepatic cell necrosis was found at the highest dose level. In the study by Shelanski of NPE with 9 ethylene oxide units, the 2 lightest male and female rats in each group were sacrificed after 8 weeks. Gross examination was made. On animals from the highest dose group, histopathological examination of 19 tissues was made. After 90 days, all remaining rats were sacrificed and subjected to macroscopic examination. Nine organs were weighed, and histopathological examination was made of 19 tissues from the 2 lightest males and females in each group. In the highest dose group, 11 of 15 rats died during the study. At 0.64% (300 mg/kg/day) and more in the diet, weight gain was retarded. In the two highest dose groups, rats were emaciated. This was judged to be referable to poor palatability by the authors. Food intake, however, was not significantly lower. This apparent contradiction was not discussed by the authors. No histopathological changes indicating toxic effects were seen. In the Dow studies of NPE with 9 or 40 ethylene oxide units, haematocrit, white blood cell total and differential counts, and haemoglobin were determined on 5 females of the control and the two highest dose levels prior to sacrifice. Six organs were weighed and 8 organs were studied histopathologically. For NPE with 9 ethylene oxide units, at the lowest dose level, 0.1% (100 mg/kg/day), no effects at all were found. At 0.3% (200 mg/kg/day) and above, relative liver weights were high. At the highest dose level, 1.0% (900 mg/kg/day), relative kidney and spleen weights were high, and growth was reduced. In the liver, petecchial areas of central lobular granular degeneration and necrosis were seen. For NPE with 40 ethylene oxide units, no effects were found at or below 0.3% (200 mg/kg/day). At 1% (700 mg/kg/day)and above relative liver weights of male rats were non-significantly heavier than controls, with general cloudy swelling and slight central lobular granular degeneration and necrosis at 3% (2000 mg/kg/day). Table 5. Results of 90-day feeding studies in rats with nonyl phenol ethoxylates of various ethylene oxide (EO) chain lengths (compiled from Smyth & Calandra 1969)
1) Low kidney weight in females. 2) Cloudy swelling of central hepatic cords, intracellular lipoid, reduced polysaccharide. 3) Focal hepatic-cell necrosis, intracellular lipoid, reduced polysaccharide. 4) Slight petecchial areas of central lobular granular degeneration and necrosis. 5) Increased relative spleen weight in females & increased relative kidney weight in males. 6) Slight central lobular granular degeneration and necrosis with general cloudy swelling. n.d.) not done (only the highest dose group was examined histopathologically) *) In the original report, dose was given as % in diet, and was subsequently calculated in mg/kg/day, based on food consumption data, by the rapporteur. NPE toxicity in dogs In the same report (Smyth & Calandra 1969), investigations of NPE toxicity in dogs was described (Table 6). Each dosage of NPE with 4, 6, 15, 20 and 30 ethylene oxide units was administered orally in gelatine capsules to 2 male and 2 female dogs. NPE with 9 ethylene oxide units was administered in the diet, a single dog per dose, and 3 dogs in the control group.
1) Focal myocardial necrosis or degeneration (microscopically). 2) Death, grossly detectable focal myocardial necrosis, lung hyperaemia. Cardiotoxicity in the dog A number of exploratory/confirmatory experiments, with cardiotoxicity as the endpoint of interest, reported by Smyth & Calandra (1969), are presented in Table 7. The details of these studies are not well-described in the publication.
Cardiotoxicity in other species Three cats and 3 rabbits were dosed by gavage with 0.7 g/kg of NPE with 20 ethylene
oxide units daily for 14 days. These two species did not develop focal myocardial necrosis
(Smyth & Calandra 1969). 2-year oral administration NPEs with 4 and 9 ethoxylene oxide units have been administered orally to rats and dogs over periods of 2 years (Smyth & Calandra 1969). Results are shown in Table 8. rats Groups of 35 male and 35 female Sprague-Dawley rats, with 5 male and 5 female rats in addition in the highest and control groups received NPE with 4 ethoxylene oxide units at 3 dose levels plus control. After 12 months 5 rats of each sex from the highest dose and control group; and 3 of each sex from the two lower-dosage groups were sacrificed. After 24 months all rats were sacrificed. Livers, kidneys, and testes were weighed. Histopathological examination of 28 tissues was done on 5 rats of each sex in the highest dose and control group. At 200 mg/kg/day females showed a reduced weight gain after 12 months, but not after 24 months. At 1000 mg/kg/day, this effect was also found in male rats. NPE with 9 ethoxylene oxide units was administered to 3 dose groups plus a control group of 36 male and 36 female Carworth-Elias rats for 2 years. Sixteen rats of each sex were interim sacrificed, at 6 and 12 months. At sacrifice livers and kidneys were weighed, and 11 tissues were histopathologically examined. There was no difference between control and treated rats in any observation made. No increased frequency of tumours was reported in either rat study. dogs NPE with 4 ethoxylene oxide units was administered for 2 years to groups of 3 male and 3 female Beagle dogs in gelatine capsules in dosages 40, 200, and 1000 mg/kg/day. An untreated control group was present.. Haematology and blood and urinary clinical-chemical parameters were measured repeatedly during the study. At sacrifice, liver, kidneys, spleen, heart, brain and testes were recorded, and 28 tissues were examined microscopically. At 200 mg/kg/day, and more pronounced at 1000 mg/kg/day, there was a moderate elevation in serum alkaline phosphatase and relative liver weight, without histopathological changes. NPE with 9 ethoxylene oxide units was administered in the diet in concentrations of 0, 0.03, 0.09, and 0.27%, corresponding to 0, 8.5, 28, and 88 mg/kg/day (author’s calculation), to groups of 3 male and 3 female Beagle dogs for 2 years. Haematology and blood and urinary clinical-chemical parameters were measured repeatedly. At sacrifice liver, kidneys, and heart was weighed, and 21 tissues were examined histopathologically. NPE with 9 ethoxylene oxide units caused an increased relative liver weight at 0.27% in the diet (88 mg/kg/day) without accompanying histopathological findings. .
Dermal contact No data have been found. 4.3 Reproductive / developmental effects Oestrogenic effects of NP and NPE Some alkyl phenols have been implicated in the hypothesis that low-level exposure can disrupt the human endocrine system, that is, that alkyl phenols may act as endocrine disrupters. Alkyl phenols, including NP, have been shown in laboratory studies to mimic the effects of oestrogen in vitro and in vivo (Lee & Lee 1996, Odum et al. 1997). The oestrogenic effect of nonylphenol and nonylphenol ethoxylates in fish and Daphnids has been studied by a number of authors. Generally the work shows that nonylphenol and nonylphenol ethoxylates do exhibit oestrogenic activity. For nonylphenol ethoxylates the activity was found to increase with decreasing chain length, with nonylphenol showing the greatest activity. (EU-RAR 1998). The oestrogenic activity of nonylphenol has been investigated in a number of studies using either recombinant yeast, oestrogen sensitive human breast tumour MCF-7 cells, or a rodent uterotrophic assay response. None of these assays have been validated as an internationally accepted toxicity test method, although the MCF-7 and uterotrophic assays have been established for a number of years as standard assays for oestrogenic activity. It should be noted that the significance to human health of oestrogenic activity detected in these assays has yet to be established. (EU-RAR 1998). NP effects in rodent uterotrophic assay The oestrogenic activity of nonylphenol in mammals has been assessed in several studies using an assay based upon the uterotrophic response in the rat. Although not stated, it is assumed that the studies have been performed with commercial grade NP which is the branched type. In the first study, five groups of immature (aged 20 - 22 days) female rats (six in each group) of a Wistar derived strain received single oral gavage doses of nonylphenol in corn oil on each of three consecutive days (ICI 1996 - quoted in EU-RAR 1998). The dose levels ranged from 9.5 to 285 mg/kg/day. Vehicle and positive (oestradiol benzoate 8 µg/kg, by subcutaneous route) groups were included. One day after the final dose the females were killed and the uterus was removed from each animal and weighed. Absolute uterus weight and bodyweight related uterus weight were statistically significantly increased, in a dose-dependent manner, at levels of 47.5 mg/kg/day and above. The NOAEL was 9.5 mg/kg/day. The uterine response seen in the positive control group was much greater than that of the nonylphenol groups, although a direct comparison of potency is not possible given the differing exposure routes. Similar data from the same laboratory have also been presented in peer-review literature (Odum et al. 1997). This latter report also included oral positive control groups (17ß-oestradiol, 10-400 µg/kg), which indicated that oestradiol was about 1000 times more potent in this assay than nonylphenol. In a similar assay, groups of ten ovariectomised female Sprague-Dawley rats were dosed once daily for three consecutive days with ethanol/oil suspensions of nonylphenol at levels of 0 (vehicle control), 30, 100 and 300 mg/kg/day (Chemical Manufacturers Association 1997b - quoted in EU-RAR 1998). Positive control groups received ethynyloestradiol in ethanol at levels of 10, 30 and 80 µg/kg/day according to the same dosing regimen. The route of administration was not stated. One day after the final dose the females were killed and the uterus was removed from each animal and weighed. Uterus weights at 300 mg/kg/day were significantly increased (1.5-fold) in comparison with the vehicle control group. A slightly greater response (a 2-fold increase) was seen in the 30 and 80 µg/kg/day positive control groups. In another uterotrophic assay, groups of three immature (aged 20-21 days) Sprague-Dawley rats each received a single intraperitoneal injection of nonylphenol at dose levels of 0, 1, 2 or 4 mg/animal (approximately 25, 50 or 100 mg/kg) (Lee and Lee 1996). Oestradiol, administered by the same route, served as a positive control. The animals were killed 24 hours later and each uterus was removed, weighed and analysed for protein and DNA content and peroxidase (thought to be a uterotrophic marker enzyme) activity. There was a dose-dependent and statistically significantly increase in uterine weight at all levels, with associated increases in uterine protein and DNA content and uterine peroxidase activity. In further experiments, the uterotrophic activity of nonylphenol was found to be blocked by the co-administration ICI 182,780, an oestrogen antagonist, providing evidence that the effect of nonylphenol is mediated through the oestrogen receptor. Also, the potency was compared with oestradiol; in this assay oestradiol was found to be about 1000 - 2000 times more potent than nonylphenol. Overall, these in vitro and in vivo studies show that nonylphenol has oestrogenic activity of a potency that is between 3 to 6 orders of magnitude less than that of oestradiol. The oestrogenic effect of NP appears to be mediated through the oestrogen receptor since its action can be blocked by oestrogen antagonists. The structure of the aliphatic side-chain (nonyl) is believed to be highly important for oestrogenic activity. Linear NP is not oestrogenic, while one or more of the branched-chain isomers may be able to mimic oestradiol in binding to the oestrogen receptor. (Odum et al. 1997). NP oestrogenic effect in 28- and 90-day study In the 28-day and 90-day studies of NP described in section 4.2 sexual organ morphology, oestrous cycle pattern, and spermatogenesis were not found to be affected by treatment at any dose level. In the 90-day study the absolute, but not relative, ovary weight in the highest dose group was slightly decreased without accompanying histopathological changes. Niehs three-generation study of NP The effects of nonylphenol on fertility and reproductive performance have been investigated in a comprehensive, good-quality multigeneration study, conducted in compliance with GLP (NIEHS 1998 - quoted in EU-RAR 1998). The overall study design was based on the OECD two-generation reproduction toxicity study guideline, with an extension to include the production of an F3 generation. This study has previously been described in the present report in relation to long term toxicity. Groups of thirty male and thirty female Sprague-Dawley rats were exposed to nonylphenol via incorporation in the diet at concentrations of 0 (control) 200, 650 or 2000 ppm over three generations. Calculated nonylphenol intakes were, respectively, about 0, 15, 50 and 160 mg/kg/day during non-reproductive phases and rising to around 0, 30, 100 and 300 mg/kg/day during lactation. Nonylphenol exposure commenced for the F0 generation at about 7 weeks of age and continued until study termination when the F3 generation were about 8 weeks old. F0 animals were mated (one male with one female) within each dose group to produce the F1 generation, selected F1 animals were similarly mated to produce the F2 generation and selected F2 animals were mated to produce the F3 generation. For the F0 generation and retained F1, F2 and F3 animals, clinical signs of toxicity, bodyweights and food consumption were reported. Oestrous cycles were monitored prior to mating. At the necropsy of adult animals, sperm samples were taken (but not from the F3 generation) for analysis of density, motility (using a computer assisted sperm motion analysis system, only conducted on control and high dose group males) and morphology, a number of organs were weighed and selected organs were sampled for histopathology. Additionally, testicular spermatid counts were made. Parameters assessed in the young offspring included litter size, bodyweights, survival, gross appearance, ano-genital distance, sexual development and, for animals killed at weaning, gross appearance of organs at necropsy and reproductive organ weights. There was evidence of general toxicity in adults of all generations, seen as a reduction in bodyweight gain at 50 and 160 mg/kg/day and histopathological changes in the kidneys at all dose levels. Considering the reproduction-related parameters, there were no adverse effects on fertility or mating performance. However, several other parameters were affected. Oestrous cycle length was increased by about 15% in the F1 and F2 females at 160 mg/kg/day, in comparison with controls. The timing of vaginal opening was accelerated by 1.5-7 days at 50 mg/kg/day and by 3-6 days at 160 mg/kg/day in females of the F1, F2 and F3 generations. Also, absolute ovarian weights were decreased at 50 mg/kg/day in the F2 generation and at 160 mg/kg/day in the F1, F2 and F3 generations; however, no effect on ovarian weight was apparent in the F1 and F3 generations when analysed as an organ-to-bodyweight ratio. In males, changes in sperm endpoints were seen only in the F2 generation; epididymal sperm density was decreased by about 10% at 50 and 160 mg/kg/day and spermatid count was decreased by a similar amount at 160 mg/kg/day. However, there may have been methodological problems with the epididymal sperm density measurements, because the density in all F2 generation groups, including controls, was considerably greater (by about 25-40%) than reported for the F0 and F1 generation males; the age of each generation was similar at necropsy, so major differences in the sperm density would not be expected. NP effects on the foetus In a study performed according to OECD guideline 414, of GLP quality, female rats were administered corn oil solutions of NP (presumably by gavage) at dose levels of 0, 75, 150, and 300 mg/kg/day from gestational day 6 to 15. Females were killed on day 20 of gestation, and foetuses were examined. At the highest and second-highest dose levels, maternal toxicity was evident, with mortality of two females. No maternal toxicity was found at 75 mg/kg/day. Post-implantation loss, litter size, foetal weights, and incidence of foetal abnormalities was not affected, even at dose levels which caused maternal toxicity (EU-RAR 1998). NPE effects on the foetus Meyer et al. (1988) administered NPE with 9 or 30 ethylene oxide units (NPE 9, NPE 30) to pregnant rats from gestational day 6 to 15. Doses of NPE 9 were 50, 250, or 500 mg/kg/day by gavage. A satellite group received 500 mg/kg/day by gavage on gestational day 1-20, and two further groups received 50 or 500 mg/kg/day dermally on gestational day 6-15. NPE 30 was given at 50, 250, or 1000 mg/kg/day by gavage on gestational day 6-15, with a satellite group receiving 1000 mg/kg/day on gestational day 1-20. The rats were killed on day 21 and the foetuses examined. NPE 30 did not cause any toxic effects on dams or foetuses. NPE 9, at 500 mg/kg, caused an increase in extra ribs in the foetuses. This dose caused a decreased weight gain in the dams. 4.4 Genotoxic effects Nonylphenol NP appears to be negative in the Ames test (Salmonella typhimurium), however, a study of sufficient quality has not been found (EU-RAR 1998). In an in vitro mammalian cell gene mutation test performed according to the OECD test guideline 476, and of GLP quality, and confirmed by a second independent experiment, NP was found non-mutagenic. The dose level may not have been high enough (EU-RAR 1998). Nonylphenol ethoxylate An NPE with unknown ethylene oxide chain length was found negative in the Ames test, using Salmonella typhimurium (CIR 1983). NPEs with 9 or 30 ethylene oxide units have been found negative in the Ames test (Salmonella typhimurium) (Meyer et al. 1988). 4.5 Carcinogenic effects For nonylphenol, no data have been found (EU-RAR 1998). For NPE with 9 ethylene oxide units the group size was 36 of each sex, with 16 rats of each sex interim sacrificed at 6 and 12 months. Histopathology was performed on 11 tissues and all neoplasms. No increased frequency of tumours was reported in either study. 5. Regulations, limit values
6. SummaryDescription NP is a clear to pale yellow viscous liquid with a slight phenolic odour. Environment NP and NPEs do not occur as natural substances. NP is released to the environment from production and use as a chemical intermediate, in the polymer industry, and as nonylphenol itself. The major part is released to water, while small percentages are released to air and soil. NP is thought to be degraded in air by reaction with photochemically produced hydroxyl radicals. Human exposure The routes of human exposure to NP and NPEs are dermal contact and inhalation by workers involved in manufacture and use, dermal and inhalation exposure of consumers from household pesticide products, dermal contact to cleaning products and cosmetics, mucous membrane contact to spermicides; and exposure via the environment through drinking water, air, and food. Toxicokinetics After oral administration, absorption of NP in humans may be significant. Limited data indicate that first pass metabolism may account for inactivation by conjugation of a major part of the NP present in blood. The main part of NP and NPE metabolites are believed to be glucuronic acid conjugates, which are excreted via the kidney. No data exist regarding inhalation or dermal exposure, but it is assumed that absorption will occur. Human toxicity There are no data on human oral or inhalatory toxicity of NP or NPE. Animal toxicity single exposure The LC50 of NP and NPE by inhalation is unknown. For NP, no data have been found. NPEs with 4 or 7 ethylene oxide units caused no mortality in concentrations up to 20 mg/l, indicating a low acute toxicity of NPE by inhalation. Oral LD50 values for NP mostly in the range of 1200 - 2400 mg/kg have been reported indicating a low order of acute oral toxicity of NP. For various NPEs (number of ethylene oxide units 4-10, 12, 13, 15 or 20) oral LD50 values between 1000 and more than 5000 mg/kg have been reported. Acute oral toxicity seems to be independent of NPE ethylene oxide chain length. A dermal LD50 of 2031 mg/kg NP in rabbits has been reported. The acute dermal toxicity of NPEs with 4, 5, 7, 9,10, 12, 13 and 40 ethylene oxide units is low, as LD50 values varying between 2000 mg/kg and 5-10,000 mg/kg have been reported. local effects The sensory irritation potential of nonylphenol has been investigated in mice in a respiratory depression test. At 3636 mg/m3 a mean respiratory rate depression of about 25% was found during exposure. However, at 267 mg/m3 there were no changes in the respiratory rate. These results suggest that nonylphenol can cause mild sensory irritation to the respiratory tract at high exposure levels. NP is corrosive on contact with skin and is a severe eye irritant. Exposure to the saturated vapour may lead to mild sensory irritation of the respiratory tract. repeated exposure In a 28-day study performed according to OECD guideline 407, rats received doses of 0,
25, 100 or 400 mg/kg b.w./day of NP in the diet. In a 90-day study performed according to
EPA guidelines and of GLP quality, rats were administered NP in the diet at levels of 0,
15, 50 and 150 mg/kg b.w./day. In both studies reduced body weight, food consumption, and
food utilisation was found; while relative kidney and liver weights were increased. In the
90-day study, females showed a slightly decreased ovary weight. In both studies, these
biologically and statistically significant adverse effects were found at the highest dose
level only. The NOAEL of the 28-day study is thus considered to be 100 mg/kg b.w./day by
the rapporteur, while the NOAEL in the 90-day study is considered to be 50 mg/kg b.w./day
by the rapporteur. NPEs with 4 or 9 ethoxylene oxide units have been administered orally to rats and dogs over periods of 2 years in studies of non-guideline standard. In the dog study, the group size was only 3 of each sex. In rats, reduced weight gain was observed after administration of NPE with 4 EO. In dogs, increased relative liver weight without accompanying histopathological findings, but with elevated serum alkaline phosphatase level was observed. In both rats and dogs, the NOAEL for NPE with 4 ethoxylene oxide units was 40 mg/kg/day. NPE with 9 ethoxylene oxide units did not cause any effects in rats in highest dose administered, 270 mg/kg/day. In dogs, the only effect was an increased relative liver weight without accompanying histopathological findings at the highest dose level, 88 mg/kg/day. Reproductive and developmental effects NP has been shown to mimic the effects of oestrogen via activation of the oestrogen receptor. NP has shown oestrogenic effect in fish, daphnids, and in human breast tumour cells. In uterotrophic assays NP has shown oestrogenic effect in immature Wistar rats, in immature Sprague-Dawley rats, and in ovariectomised Sprague-Dawley rats. In 28-day and 90-day studies of NP, the latter study having special emphasis on reproductive organ function and morphology, sexual organ morphology, oestrous cycle pattern, and spermatogenesis were not found to be affected by treatment. Absolute, but not relative, ovary weight was slightly reduced in the highest dose group in the 90-day study. A three-generation study has revealed effects on female and male reproductive parameters of NP, without effects on fertility. The NOAEL for reproductive effects in this study was 15 mg/kg/day. Neither NP nor NPE (9 or 30 ethylene oxide units) induced malformations in rat foetuses exposed during organogenesis. Genotoxicity The existing mutagenicity studies of NP are not of sufficient quality to allow a proper evaluation of mutagenic potential. NPEs with 9 or 30 ethylene oxide units were negative in the Ames test. Carcinogenicity For NP, no data have been found. NPEs with 4 or 9 ethylene oxide units have been administered in 2-year studies. However, because of an insufficient number of animals, the studies do not allow a proper evaluation of carcinogenic potential. 7. EvaluationA study in humans involving two volunteers provides evidence that nonylphenol is rapidly absorbed from the gastrointestinal tract in humans. Also, the fact that only a small proportion of the dose was recovered in faeces within 56 hours suggests that almost complete absorption of a dose had occurred. Following oral administration, most of the nonylphenol was present in the blood as glucuronide or sulphate conjugates, in contrast to the findings for intravenous administration where similar proportions of unconjugated and conjugated nonylphenol were detected; these findings are indicative of extensive first pass metabolism. First pass metabolism does not occur when the route of exposure is inhalation. Therefore, an inhalatory dose of NP may be more toxic than the same dose administered orally. No data have been found which describe the health effects of NP or NPEs in humans after oral ingestion or exposure via inhalation. The systemic toxicity of NP and NPE by inhalation is unknown. Data from a mouse study suggest that NP can cause mild sensory irritation to the respiratory tract at high exposure levels. At a vapour concentration of 267 mg/m3 there was no effect. The acute oral toxicity of NP and NPE is low. NP has not been tested for mutagenicity in a study of good quality. No data regarding the carcinogenicity of NP have been found. NPEs with 9 or 30 ethylene oxide units were negative in the Ames test. Two-year feeding studies showed no tumourigenic effects of NPEs with 4 or 9 ethylene oxide units, however, the studies did not include a sufficient number of test animals and are for this reason inconclusive. Nonylphenol A 28-day study NOAEL of 100 mg/kg b.w./day for NP, and a 90-day study NOAEL of 50 mg/kg b.w./day for NP have been determined in studies of good quality. The results of the 28-day study and the 90-day study are in agreement with each other. In a multigeneration study, also of good quality, increased incidence of renal tubular degeneration and/or dilatation were found. There was no dose level without effect. It is difficult to decide for certain whether or not the kidney effect was related to treatment because these changes were not seen in the 90-day study, which was conducted using the same strain of rats, and because a dose-dependent trend was not apparent in all generations/sexes. The lack of concordance between the studies cannot be explained on the basis of a slightly longer exposure period in the multigeneration study because kidney effects were seen in the F3 generation which were exposed for only 8 weeks, nor solely on the basis of in utero and neonatal exposure because the effect also occurred in the F0 generation. Giving special emphasis to the fact that the increased incidence occurred consistently across all four generations in the multigeneration study, it is considered that this cannot be dismissed as background variation. The F3 generation showed the highest incidence of kidney changes, indicating that the effect may become more pronounced after exposure during several generations. Consequently, the conclusion has been drawn from this study that there is a LOEL for repeated exposure of 15 mg/kg/day, based on histopathological changes in the kidneys. Since renal tubular degeneration and/or dilatation are common findings in untreated rats, and as they were not accompanied by other related signs or symptoms in the affected rats, they are not considered signs of severe toxicity by the rapporteur. Alkyl phenols, including NP, have been shown in laboratory studies to mimic the effects
of oestrogen in vitro and in vivo. However, the results of the three-generation study indicate that effects on male and female reproductive parameters may occur at dose levels above 15 mg/kg/day. In females, accelerated sexual maturation, increased oestrous cycle length and reduced ovarian weights were found, while males exhibited a reduced number of spermatids. The effects were found in one or more of the three filial generations. NOAEL for NP In conclusion, for NP a LOAEL of 15 mg/kg/day is set, based on the reproductive/developmental effects seen in the oral three-generation study. This dose level is also a LOEL for the kidney effects also identified in this study. Nonylphenol ethoxylate For NPEs, a number of 90-day studies in rats and dogs exist. The studies are all rather
old (reports dated 1959-1965) and probably do not fulfil present standards. Histopathology
apparently has only been performed in the highest dose groups, and therefore it is not
possible to evaluate the toxicological significance of increased organ weights of other
dose groups. For this reason, a change in organ weight was regarded as an adverse effect
in the absence of histopathological data. NOAELs ranging between 40 and 160 mg/kg/day in
rats for NPEs with 4-15 ethylene oxide units have been found. In dogs, 90-day NOELs of 40
mg/kg/day were found for NPEs with 4, 6, 9, 15, and 20 ethylene oxide units. For NPEs with
ethylene oxide chains of 20 or 30, NOAELs of 1000 ->5000 mg/kg/day were found in rats.
In dogs, the NOAEL for NPE with 30 ethylene oxide units was >1000 mg/kg/day. NPE with
40 ethylene oxide units had a NOAEL of 300 mg/kg/day in rats. NPE effects included
retarded growth and increased liver weight, for some compounds necrosis of liver cells. NPEs with 4 or 9 ethoxylene oxide units have been administered orally to rats and dogs for periods of 2 years. Effects included reduced weight gain, and increased relative liver weight, while no increased frequency of tumours was reported. However, the studies were not properly designed to evaluate carcinogenic effects. In the dog studies, the group size was very small, and the 2-year dose period did not cover the whole lifetime for this species, which is at least 7-8 years. In the rat studies, the number of animals examined in detail for tumour occurrence was far too small to allow any conclusions. In rats and dogs, the 2-year chronic toxicity NOAEL for NPE with 4 ethoxylene oxide units was 40 mg/kg/day, while NPE with 9 ethoxylene oxide units did not cause any effects in rats in highest dose administered, 270 mg/kg/day. In dogs, the 2-year NOAEL for NPE with 9 ethoxylene oxide units was 88 mg/kg/day. NOAEL for NPE In conclusion, an oral NOAEL of 40 mg/kg/day for NPEs with ethylene oxide chain lengths shorter than 15 and between 21-40 can be set. This NOAEL covers the most toxic NPEs among the group tested (Table 6). For NPEs with ethylene chain lengths between 15 and 20 a NOAEL cannot be determined from the available data. The LOAEL for cardiotoxicity in the dog is 40 mg/kg For the purpose of setting a limit value covering all NPEs a LOAEL of 40 mg/kg/day is set. 8. TDI, health based limit values8. TDI, health based limit values 8.1 TDI NP = 0.005 mg/kg b.w./day The safety factor SFI is set to 10 assuming that humans are more sensitive than animals. The SFII is set to 10 to protect the most sensitive individuals in the population. The SFIII is set to 30 since a LOAEL is used and because data for genotoxicity and carcinogenicity are lacking. NPE = 0.013 mg/kg b.w./day The safety factor SFI is set to 10 assuming that humans are more sensitive than animals. The SFII is set to 10 to protect the most sensitive individuals in the population. The SFIII is set to 30 since a LOAEL is used and because data for genotoxicity and carcinogenicity are lacking. Allocation The sources of NP and NPE exposure for the general population are via consumer products and via the environment through food, drinking water, and air. The size of these exposures is unknown, although it is expected that exposure via air is negligible. For the purpose of setting a limit value, only 10 % of the TDI is allocated to ingestion of soil and 10 % to drinking water. 8.2 Limit value in soil NP Based on the TDI of 0.005 mg/kg b.w. per day for NP, and 0.04 mg/kg b.w. per day for NPE, and assuming a daily ingestion of 0.2 g soil for a child weighing 10 kg (wchild), a limit value is calculated: = 25 mg/kg soil NPE Based on the TDI of 0.013 mg/kg b.w. per day for NPE, and assuming a daily ingestion of 0.2 g soil for a child weighing 10 kg (wchild), a limit value is calculated: = 65 mg/kg soil 8.3 Limit value in drinking water NP and NPE The existing limit value for phenols in drinking water of 0.5 m g/l offers adequate protection against adverse health effects induced by NP or NPE. 8.4 Limit value in air NP = 0.017 mg/m3 NPE = 0.05 mg/m3 9. Quality criteria9. C-value 9.1 Quality criteria in soil NP A limit value of 25 mg/kg has been calculated based on children’s ingestion of soil. A quality criterion of 25 mg/kg soil is proposed. Quality criteria Quality criterion: 25 mg/kg soil NPE A limit value of 65 mg/kg has been calculated based on children’s ingestion of soil. A quality criterion of 65 mg/kg soil is proposed. Quality criteria Quality criterion: 65 mg/kg soil 9.2 Quality criteria in drinking water The existing limit value for phenols in drinking water of 0.5 m g/l offers adequate protection against adverse health effects induced by NP or NPE. Quality criteria Quality criterion: 0.5 g m/l. 9.3 C-value NP A limit value of 0.017 mg/m3 has been calculated based on toxicological considerations in relation to repeated oral exposure. For substances having acute or subchronic effects, but for which activity over a certain period of time is necessary before the harmful effect occurs, the C-value is set at the limit value. A C-value of 0.02 mg/m3 and placing in Main Group 1 is proposed. Main Group 1 is justified because of concern for the environment due to the persistence of NP. This proposed C-value also protects against sensory irritation caused by NP. C-value 0.02 mg/m3, Main Group 1. NPE A limit value of 0.05 mg/m3 have been calculated based on toxicological considerations in relation to repeated oral exposure. For substances having acute or subchronic effects, but for which activity over a certain period of time is necessary before the harmful effect occurs, the C-value is set at the limit value. A C-value of 0.05 mg/m3 and placing in Main Group 1 is proposed. Main Group 1 is justified because of concern for the environment due to the persistence of NP. C-value 0.05 mg/m3, Main Group 1 10. ReferencesAnonymous (1997). APE producers offering data, co-operation. INFORM 12, 1271-1279. CIR (Cosmetic Ingredient Review (1983). Final report on the safety assessment of nonoxynols -2, -4, -8, -9, -10, -12, -14, -15, -30, -40, and -50. J Am Coll Toxicol 2, 35-60. Cunny HC (1997). Subchronic toxicity (90-day study) with para-nonylphenol in rats. Reg Toxicol Pharmacol 26, 172-178. EU-RAR (1998). Risk assessment draft, September 1998 (version for TMIII98). HSDB (1997). Polyethylene glycol nonylphenyl ether. In: Hazardous Substances Data Base. IPCS (1998). International programme on chemical safety. Environmental health criteria. Nonylphenol and nonylphenol ethoxylates. First draft - June 1998. Knaak JB, Eldridge JM and Sullivan LJ (1966). Excretion of certain polyethylene glycol ether adducts of nonylphenol by the rat. Toxicol Appl Pharmacol 9, 331-340. Lee P-C and Lee W (1996). In vivo estrogenic action of nonylphenol in immature female rats. Bull Environ Contam Toxicol 57, 341-348. Meyer O, Andersen PH, Hansen EV and Larsen JC (1988). Teratogenicity and in vitro mutagenicity studies on nonoxynol-9 and 30. Pharmacol Toxicol 62, 236-238. MM (1988). Bekendtgørelse om vandkvalitet og tilsyn med vandforsyningsanlæg. Miljøministeriets bekendtgørelse nr. 515 af 29. august 1988. MST (1990). Begrænsning af luftforurening fra virksomheder. Vejledning fra Miljøstyrelsen nr. 6 1990. MM (1996). Bekendtgørelse om anvendelse af affaldsprodukter til jordbrugsformål. Nr. 823 af 16. september, 1996. MST (1996). B-værdier. Orientering fra Miljøstyrelsen Nr. 15 1996. MM (1997). The Statutory Order from the Ministry of the Environment no. 829 of November 6, 1997, on the List of Chemical Substances. NIEHS (1998). Nonylphenol: Multigenerational reproductive effects in Sprague-Dawley rats when exposed to nonylphenol in the diet (CAS 84852-15-3). Study number: RACB94021. (Accompanied by data tables). Odum J, Lefevre PA, Tittensor S, Paton D, Routledge EJ, Beresford NA, Sumpter JP and Ashby J (1997). The rodent uterotrophic assay: critical protocol features, studies with nonyl phenols, and comparison with a yeast estrogenicity assay. Reg Toxicol Pharmacol 25, 176-188. Smyth HF and Calandra JC (1969). Toxicological studies of alkylphenol polyoxyethylene surfactants. Toxicol Appl Pharmacol 14, 315-334. Superfos Kemi a/s (1997).‘Synperonic’ NP surfactants (Enclosure to letter from Superfos Kemi a/s to Miljøstyrelsen, 29 October 1997). Swisher RD (1970). Surfactant biodegradation. Marcel Dekker, Inc., New York Talmage SS (1994). Environmental and human safety of major surfactants. Alcohol ethoxylates and alkylphenol ethoxylates. The soap and detergent association. Lewis Publishers. Boca Raton, Ann Arbor, London, Tokyo. May 1999 / final Evaluation of health hazards by exposure to Tricresyl phosphatesand estimation of quality criteria in soil and drinking water. Jens Erik Jelnes Ole Ladefoged Elsa Nielsen The Institute of Food Safety and Toxicology Danish Veterinary and Food Administration 1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 7. Evaluation 8. TDI, health based limit values 9. C-value 1. General description1. General description Tricresyl phosphate exists as 10 different pure isomeric substances with the three cresyl groups being either ortho, meta, or para. The name tricresyl phosphate is used as a common name for the ten substances and for mixtures of these. The relative content of the different cresyl moieties depends on the cresols used in the production of tricresyl phosphate. Industrial-grade tricresyl phosphate contains predominately the meta- and para-isomers and modern mixtures contain less than 1% of the ortho-isomer. This document will deal with both the mixed tricresyl phosphates and with the pure tri-ortho, tri-meta, and tri-para-cresyl phosphates. In addition, studies, where tricresyl phosphate is the major constituent, are included, as long as they are sold / named tricresyl phosphate e.g. studies with mixtures where cresyl-xylenyl or cresyl-phenyl phosphates occurs. 1.1 Identity
2) 3) 4)
1.2 Physical / chemical properties
1.3 Production and use The world production of TCP in unknown but Japan produced 33000 tonnes in 1984, USA 10400 tonnes in 1977, and China produced about 1000 tonnes in 1989 (EHC 1990). TCP is produced by the reaction of cresols with phosphorous oxychloride. The cresols can be derived from cresylic acid or tar acid, which is a mixture of isomers of cresol and varying amounts of xylenols, phenol, and other high-boiling phenolic fractions obtained as a residue from coke ovens and petroleum refining. Using this source of cresol in the synthesis yields a very heterogeneous TCP. Another source of cresol is synthetic cresol, prepared from cymene via oxidation and catalytic degradation. This process can after purification yield ortho-, meta-, and para-cresol of high purity, which can be used to synthesise the pure tri-o-, tri-m-, and tri-p-cresyl phosphates. Mixing pure meta- and para-cresol together with phosphorous oxychloride can give TCP with variable isomer composition e.g. different proportions of the four possible combinations of meta- and para-cresyl phosphates. (EHC 1990). TCP has many uses. It has been used since the start of the century in hydraulic oils. Other main uses are as a plastisiser in vinyl plastic manufacture, as a flame-retardant, a solvent for nitrocellulose, in cellulosic molding compositions, and as an additive to extreme pressure lubricants. Minor uses are in cutting oils, machine oils, transmission fluids, and cooling lubricants. Other minor uses are as an additive in making synthetic leather, shoes, polyvinyl acetate products, as solvent for acrylate lacquers and varnishes, and in non-smudge carbon paper. (Various authors quoted from EHC 1990). 1.4 Environmental occurrenceTCP does not occur naturally in the environment. TCP is released to air, water, soil and sediment as a result of its production, processing and use. The majority of TCP release to the environment is accounted for by end-point use (particularly volatilisation from plastics and leaking hydraulic fluids) rather than production. (EHC 1990, HSDB 1998). Air No data are available on the release of TCP to the atmosphere from production processes. However, open, high-temperature processes such as roll milling, calandering and extrusion of plasticised polymers may result in significant gaseous emissions of aryl phosphates including TCP (Boethling & Cooper 1985 - quoted from EHC 1990). TCP levels of 0.01-2 ng/m3 in air collected at production sites in USA have been reported (MRI 1979 - quoted from EHC 1990). Near heavily industrialised cities in Japan TCP levels of 11.5-21.4 ng/m3 were found in three out of four samples, and in samples of urban air TCP levels of 26.7-70.3 ng/m3 were recorded in three out of 19 samples (Yasuda 1980 - quoted from EHC 1990). Water TCP is slightly soluble in water. It has only occasionally been detected in water samples (several authors - quoted from EHC 1990). In 13 out of 84 Canadian drinking water samples, TCP was detected in concentrations of 0.3-4.3 ng/l (EHC 1990). The adsorption coefficient of TCP on marine sediment was found to be 420 (Kenmotsu et al. 1980 - quoted from EHC 1990). Soil TCP has been detected in soil at a chemical plant at a level of 1.0-4.0 mg/kg (Boethling & Cooper 1985- quoted from EHC 1990). In a Danish study TCP was detected in sewage sludge at levels of 57 to 12,000 µg/kg dry matter in 11 of 20 representative sewage treatment plants. In water extracts from the same sewage treatment plants TCP was found in 9 plants with a mean concentration of 1.50 g µ/l (range 0.15 - 3.80 µg/l). (MST, 1996). Foodstuffs As TCP has a log Poctanol/water of 5.11, and as it is found in sewage sludge, bioaccumulation of the substances can occur. Bioconcentration factors of 165-2768 have been measured for several species in laboratory tests (EHC 1990). The levels of TCP given for fish and shellfish below are indicative of this, as up to 3.3 g µ/l is found in water extracts from Danish sewage treatment plants (MST 1996). This concentration will, when distributed over a greater water volume, fall considerably. A concentration of 40 ng/g of TCP has been found in sturgeon from the Colombia River, USA (Lombardo & Ergy 1979 - quoted from EHC 1990). In fish caught near triaryl phosphate manufacturing plants 2-4 ng TCP/g were found (Muir 1984 - quoted from EHC 1990). In samples of fish and shellfish caught in the Seto Inland Sea, Japan, TCP levels of 1-19 ng/g was found in 4 out of 41 samples (Kenmotsu et al. 1981 - quoted from EHC 1990). 1.5 Environmental fateAir If released to the atmosphere, TCP will degrade in the vapour-phase by reaction with photochemically produced hydroxyl radicals with an estimated half-life of 26 hours). Physical removal of particulates from air by dry deposition (settling) and wet deposition (rainfall) can occur. (HSDB 1998). Water TCP released into water is readily adsorbed on to sediment particles, and little or none remains in solution (EHC 1990). Biodegradation is an important process in aerobic waters. Screening studies suggest that TCP will be biodegraded at moderate to rapid rates with half-lives on the order of several days or less. Biodegradation under anaerobic conditions is unclear. TCP can also degrade through aqueous hydrolysis. The neutral hydrolysis half-life at pH 7 and 25° C is about one month; the hydrolysis rate will increase as the water becomes increasingly alkaline. (HSDB 1998). Among the isomers of TCP, the ortho isomer degraded in river water slightly faster than the meta isomer and both isomers degraded faster than the para isomer (Howard & Deo 1979 - quoted from EHC 1990). Soil Biodegradation is expected to be a dominant degradation process in soil. TCP is relatively immobile in soil as it adsorbs strongly to soil and is not expected to leach. (HSDB 1998). Sediment TCP is readily biodegraded in sewage sludge with a half-life of 7.5 hours, the degradation within 24 hours being up to 99% (EHC 1990). Even though TCP’s are degraded within 5 days in sewage treatment plant sludge, the substances are often found as pollution’s in nature due to their very widespread use. Biodegradation pathway The degradation pathway for TCP most probably involves stepwise enzymatic hydrolysis to orthophosphate and phenolic moieties; the phenol would then be expected to undergo further degradation. 1.6 Human exposureNo data have been found. 2. Toxicokinetics2. Toxicokinetics 2.1 Absorption, distribution o-TCP absorption has been studied in a variety of species using oral or dermal administration, and p-TCP absorption has been studied after oral dosing in rats. No information is available on absorption following inhalation. Oral intake After a single oral dose of 770 mg 32P-o-TCP/kg b.w. to chickens, the total radioactivity in the liver increased consistently throughout 72 hours. The levels of radioactivity in the plasma were consistently lower than those in the liver. At 24 hours the plasma levels were 5% of those in the liver. The radioactivity was predominantly associated with o-TCP metabolites in the liver but with unmetabolised o-TCP in blood. (Sharma & Watanabe 1974 - quoted from EHC 1990). Incomplete absorption of p-TCP in the intestine of rats was demonstrated after a single oral dose of methyl-14C-p-TCP (7.8 or 89.6 mg/kg b.w.) in 1.5 ml dimethylsulfoxide. Much of the radioactivity was recovered in the faeces, predominantly in the form of unchanged p-TCP (Kurebayashi et al. 1985 - quoted from EHC 1990). Groups of three rats were given 2, 20, or 200 mg/kg b.w. of 14C-o-TCP, 14C -m-TCP, or 14C -p-TCP. Excretion studies (see later) showed that urinary excretion was of importance (60-70% of a dose excreted via urine) (Perry et al. 1983). These data indicate that extensive absorption of all three substances from the gastrointestinal tract occurs. Dermal contact Poor absorption of 32P-labelled o-TCP was demonstrated in a dog after a single dermal dose of 200 mg/kg b.w.. In humans, the rate of transfer (dose applied 2-4 mg 32P-o-TCP/kg b.w.) through intact human palm skin appeared to be about 100 times faster than that through the abdominal skin of the dog. This was based on urinary excretion and surface area considerations. (Hodge & Sterner 1943 - quoted from EHC 1990). After a single dose of 200 mg 32P-o-TCP/kg b.w. to the abdominal skin of the dog, the radioactivity in the blood during the first 24 hours was equivalent to an average o-TCP concentration of 80 g m/l during the period. The radioactivity was distributed throughout the visceral organs, muscle, brain, and bone. The levels of radioactivity in the tissues were in decreasing order of concentration: liver > blood > kidney > lung > muscle or spinal cord > brain or sciatic nerve. (Hodge & Sterner 1943 - quoted from EHC 1990). Another species, the cat, showed great dermal absorption. When 32P-o-TCP (50 mg/kg b.w.) was dermally applied to adult male cats, the disappearance of radioactivity from the application site was bi-exponential. In the first phase, 73% of the o-TCP disappeared within 12 hours, while in the second phase the half-life was 2 days. (Nomeir & Abou-Donia 1984, 1986b - quoted from EHC 1990). In cats given a single dermal dose of 50 mg 14C-o-TCP/kg b.w., the chemical was absorbed from the skin and subsequently distributed throughout the body. o-TCP reached its highest concentration in plasma at 12 hours, and its metabolites attained their maximum concentration between 24-48 hours. The relative residence values of unmetabolised o-TCP in various tissues, relative to plasma, were: brain 0.09, spinal cord 0.18, sciatic nerve 2.1, liver 0.44, kidney 0.55, lung 1.27. Parent o-TCP was the predominant compound in the brain, spinal cord, and the sciatic nerve, while the metabolites o-hydroxybenzoic acid and di-o-cresyl phosphate were predominant in the liver, kidney, and lung (Nomeir & Abou-Donia 1984 - quoted from EHC 1990). In contrast, when measuring total radioactivity in samples 1-10 days post exposure, highest levels were found in the bile, gall bladder, urinary bladder, kidney, and liver, with only low levels in the spinal cord and brain (Nomeir & Abou-Donia 1986b - quoted from EHC 1990). 2.2 Elimination Metabolism The metabolisms of o-tri- and p-tricresyl phosphate (figures 1 and 2) are essentially the same, starting with oxidation of one of the methyl groups followed by hydrolysis of the ester bond yielding hydroxybenzyl alcohol, which is further oxidised to hydroxybenzoic acid. For o-TCP there is a "bypass" as a ring structure can be formed after the first cresyl moiety has been split off, this results in the formation of saligenin cyclic o-tolyl phosphate. The further metabolism of the substances follows the same general picture i.e. hydroxylation of the cresol moiety, which is then split off and oxidised. No studies on the metabolism of m-TCP have been published. (NTP 1994). Excretion Groups of three rats were given 2, 20, or 200 mg/kg b.w. of 14C-radiolabelled o-TCP, m-TCP, or p-TCP by gavage. For all isomers 90-100% of the dose was excreted in urine and faeces within 3 days. Rats receiving 2 and 20 mg/kg b.w. o-TCP eliminated 90% of the dose (60-70% urine, 20-25% faeces), however, at the 200 mg/kg b.w. dose only 60% of the dose had been eliminated (45% urine, 16% faeces) by 24 hours. For m-TCP equal amounts of the 20 and 200 mg/kg b.w. dose appeared in urine and faeces (24% and 12 %, respectively) by 24 hours. By three days the major route of excretion for the 20 and 200 mg/kg b.w. doses was via the faeces, while there was an equal elimination by both routes for 2 mg/kg b.w.. The excretion route and rate of p-TCP also showed dose dependence, with 59, 36, and 17% of the dose (2, 20, and 200 mg/kg b.w., respectively) appearing in the urine and 25, 46, and 47% appearing in the faeces by 24 hours. The major route of p-TCP excretion changed from 2 mg/kg b.w. (urine) to 200 mg/kg b.w. (faecal). (Perry et al. 1983). Figure 1. Metabolism of o-TCP (From NTP 1994). Look here Figure 2. The metabolism of p-TCP (From NTP 1994). Look here 2.3 Toxicological mechanisms It is known that o-TCP and TCP isomers with minimum one o-cresyl group (see below for potency) exerts their organophosphorous induced delayed neurotoxicity (OPIDN) through the metabolite saligenin cyclic o-tolyl phosphate, which is at least five times more neurotoxic than o-TCP after oral administration to chickens (Bleiberg & Johnson 1965 - quoted from EHC 1990). The formula of saligenin cyclic o-tolyl phosphate is given below. Figure 3. Structure of the neurotoxic metabolite saligenin cyclic o-tolyl phosphate. Henschler (1959) studied the relative neurotoxicity of the 10 different isomers of tricresyl phosphates after single oral doses in hens. The most potent isomers for inducing OPIDN were the three isomers with only one ortho-cresyl group, these were equally potent. When the OPIDN toxicity of these was set to 100%, the two isomers with two ortho-cresyl groups had a relative toxicity of 50%, and the tri-ortho-cresyl phosphate had a relative toxicity of 10%. The four tricresyl phosphates without the ortho-isomer did not induce OPIDN. Whether the different m- and p-cresyl phosphate isomers are neurotoxic through another mechanism is not clear. The neurotoxicity seen in the NTP (1994) studies (see later) might be explained by small amounts of tricresyl phosphate with one or more o-cresyl rings being present, as it is only stated that the TCP used contained less than 0.1% o-TCP. 3. Human toxicity3. Human toxicity 3.1 Short term toxicity TCP with at least one o-cresyl ring produce organophosphorous induced delayed neuropathy (OPIDN) in humans. The symptoms of OPIDN are sharp, cramp like pains in the calves, and some numbness and tingling in the feet, and sometimes hands. Within a few hours or a day or two at most these pains are followed by increasing weakness of the lower limbs, and soon after the patient becomes unsteady and then unable to keep his balance. One or two weeks after the onset in the lower limbs, and while paralysis may still be progressing, weakness spreads to the hands. While some patients show complete wrist drop and total loss of power in the hands, sometimes with weakness up to the elbows, the predominant neurological abnormalities are observed in the lower limbs, where bilateral drop foot with complete loss of power from the ankle down is a common finding. (Inoue et al. 1988). Inhalation No relevant data available Oral intake In Sri Lanka, acute polyneuropathy resulting in some cases of OPIDN (numbers not specified) occurred in over 20 young women in 1977 to 1978. The cause was a special cooking oil, gingili oil, contaminated by TCP. The chemical composition of the TCP mixture was stated to be an isomeric mixture of tricresyl phosphate with little or no o-TCP. The amount of other ortho isomers was not stated. The special use of this cooking oil, only taken in an egg shell daily for two weeks after the first menstrual flow, allows a daily intake of 5-10 mg TCP/kg b.w./day to be estimated. (Senanayake & Jeyaratnam 1981). According to Staehelin (1941), a single dose of 0.15 g o-TCP can produce toxic symptoms (type not indicated). According to Inoue et al. (1988) severe neurological disturbance developed in three men after oral intake of 0.5 to 0.7 g o-TCP in a Swiss incident of o-TCP poisoning. It has been suggested that individual susceptibility varies greatly (Inoue et al. 1988). 3.2 Long term toxicity Oral intake Several incidents of TCP poisoning resulting in OPIDN, involving from a few (6) to 50,000 people have been reported in the literature following oral intake of TCP (see EHC 1990) but none of them allows estimation of the daily intake. A recent study, which allows estimation of daily intake, will be described below. This is not included in the EHC review. In 1990, in China 176 people became intoxicated by TCP, containing 20% o-TCP, from flour from the local mill. The TCP was in the machine oil used to lubricate the mill, and the lubricant leaked into the flour. In all, there were 91 cases of TCP intoxication with diagnosed clear OPIDN and 85 cases with mild nausea and weakness. Twelve persons were hospitalised and after two years 9 of the 12 were still unable to walk. Flour from the mill and the households were sampled and analysed for TCP and was found to contain 0.046 to 0.181 mg TCP/g flour. An average daily dose over 10 to 46 days was calculated to be 47.5 to 76 mg TCP. (Wang & Tao 1995). Dermal contact Among two patients, who had developed allergic contact dermatitis to adhesive bandages, and who were patch tested for the chemicals contained in the adhesive bandage, one patient responded positive to tricresyl phosphate (isomer not stated) (Norris and Storrs 1990). Occupational exposure At least five cases of OPIDN have been described after occupational exposure to o-TCP or TCP. In some of the cases, dermal contact was suggested to be the route of exposure. (EHC 1990). 3.3 Reproductive and developmental effects No relevant data have been found. 3.4 Mutagenic and genotoxic effects No data have been found. 3.5 Carcinogenic effects No data have been found. 4. Toxicity, animal data4. Toxicity, animal data 4.1 Short term toxicity Animal testing is of relevance for detecting, if a substance can induce OPIDN. There is a great difference in the sensitivity of the different species in the induction of OPIDN. The chicken and cat have been widely utilised to investigate the syndrome because their responses are similar to those of man. Rabbit, guinea pig, monkey and dog react inconsistently. Rats and mice have been reported to be relatively resistant to the severe paralysis in spite of nervous tissue damage. The preferred test organism is the domestic hen. This is used in both the acute and subchronic testing for delayed neurotoxicity of organophosphorous substances in OECD Test Guidelines 418 and 419 (OECD 1993), as experience has shown that this organism is the most relevant for extrapolating the results to human effects after organophosphorous exposure. Inhalation No data have been found. Oral administration For mixed isomers of TCP, the oral LD50-values range from >4640 to >15.800 mg/kg b.w. in the rat. In the mouse a value of 3900 mg/kg b.w. has been reported. For o-TCP, a range of LD50-values from 1160 to 8400 mg/kg b.w. has been reported for the rat, and in the rabbit a value of 3700 mg/kg b.w. has been found. In chicken, o-TCP is more acutely toxic than in the other species tested with LD50-values of 100-500 mg/kg b.w. being reported. For m-TCP, the LD50-value is reported as >3000 mg/kg b.w. in the rabbit, and >2000 mg/kg b.w. in chicken. For p-TCP, the LD50-value is reported as >3000 mg/kg b.w. in the rabbit, and >2000 mg/kg b.w. in chicken. (EHC 1990). 4.1.1 Studies on all possible TCP isomers Henschler (1959) studied the relative neurotoxicity of the 10 different isomers of tricresyl phosphates after single oral doses in hens. The most potent isomers for inducing OPIDN were the three isomers with only one ortho-cresyl group, these were equally potent. When the toxicity of these is set to 100%, then the two isomers with two ortho-cresyl groups had a relative toxicity of 50%, and the tri-ortho-cresyl phosphate had a relative toxicity of 10%. The four tricresyl phosphates without the ortho-isomer did not induce OPIDN. Information on dose ranges are not given in detail. 4.1.2 Studies on o-TCP White leghorn adult hens were given a single oral dose of 0, 300, 400, 500, 600, or 700 mg o-TCP/kg b.w.. Hens were sacrificed 14 and 21 days after dosing for neuropathological study, and other hens were sacrificed day 1 and 2 after dosing for determination of neurotoxic esterase (NTE) inhibition in brain and peripheral nerve. Symptoms of OPIDN appeared in the 400, 500, 600, and 700 mg/kg b.w. group on day 3 and in the 300 mg/kg b.w. group on day 4 in the form of weakness of the legs. By day 16 all o-TCP dosed groups had complete prostration. The inhibition of NTE in brain and peripheral nerve increased dose-dependently compared to controls from 70% in the 300 mg/kg b.w. group to 90% in the 700 mg/kg b.w. group on day 1 and 65% to 80% on day 2 for the two groups, respectively. In the neuropathological study, o-TCP-treated hens showed progressive neuropathological degenerative changes in brain tissue and in peripheral nerves. In the 300 mg/kg b.w. group the mean damage score was 20% and 38% for brain and peripheral nerve, respectively. This score increased dose-dependently to 35% and 60% in the 700 mg/kg b.w. group for brain and peripheral nerve, respectively. (Nanda & Tapswi 1995). It has been shown that a single oral dose of o-TCP in the range of 58 to 580 mg/kg b.w. induces mild to severe paralysis in the hen (Cavanagh 1954, Hine et al. 1956 - quoted from EHC 1990). Certain animal species (e.g. cats, dogs, cows, sheep and chicken) are susceptible to OPIDN-related paralysis after single oral doses of down to 100 mg o-TCP/kg b.w., whereas others (rats and mice) are less susceptible to the ataxia but susceptible to the pathological changes. (EHC 1990, Wilson et al. 1982). 4.1.3 Studies on other TCP’s. In the studies given below, the isomer composition is given, when stated in the study. Groups of 10 male and 10 female F344/N rats received TCP in corn oil by gavage at doses of 0, 360, 730, 1450, 2900, or 5800 mg/kg b.w. 5 days per week for a total of 13 or 14 doses during a 16-day period (NTP 1994). The tricresyl phosphate used in this and the subsequent NTP studies contained 21% m-TCP, 4% p-TCP and less than 0.1% o-TCP. Gas chromatographic analysis revealed two major peaks (24 and 30% of the total area) being isomers for which the chemical composition was not revealed. Dicresyl phosphate esters comprised 17% of the mixture. One female receiving 1450 mg/kg b.w./day and five males and eight females receiving 2900 mg/kg b.w./day died before the end of the study. No deaths occurred in the other dose groups including the 5800 mg/kg b.w./ day group. The only clinical finding related to TCP was diarrhoea, which occurred from 730 mg/kg b.w./day. Final mean body weights of male and female rats that received 1450, 2900, or 5800 mg/kg b.w./day were significantly lower than those of the controls. Necrosis of the mandibular lymph node, spleen, and thymus occurred primarily in rats receiving 2900 and 5800 mg/kg b.w./day. Diffuse aspermatogenesis occurred in the testes of male rats that received 2900 and 5800 mg/kg b.w./day. Changes in neurobehavioural parameters (spontaneous motor activity, forelimb and hindlimb grip strength, startle response, and paw-lick latency) were seen in groups that received 1450, 2900, or 5800 mg/kg b.w./day. The neurobehavioural results were confounded by mortality and reduced body weights and were not attributed to a direct neurotoxic response, so no detailed results were given. Groups of 10 male and 10 female B6C3F1 mice received TCP in corn oil by gavage at doses of 0, 360, 730, 1450, 2900, or 5800 mg/kg b.w. 5 days per week for a total of 13 or 14 doses during a 16-day period (NTP 1994). Five males and all females that received 1450 mg/kg b.w./day, all mice that received 2900 mg/kg b.w./day, and four males and one female that received 5800 mg/kg b.w./day died before the end of the study. Final mean body weights of males in the 1450 and 5800 mg/kg b.w./day groups were significantly lower than those of the controls. Final mean body weight of females of the 360, 730, and 5800 mg/kg b.w./day groups were significantly greater than those of the controls. Necrosis of the mandibular lymph node, thymus, and spleen occurred primarily in mice receiving 2900 and 5800 mg/kg b.w./day. Hindlimb grip strength of male mice that received 360 and 1450 mg/kg b.w./day and male and female mice that received 730 and 5800 mg/kg b.w./day were significantly lower than those of the controls at the end of the study. Male and female CD-1 mice were given a diet containing 0, 0.437, 0.875, 1.75, 3.5, or 7% TCP for 14 days. No signs of toxicity developed at doses up to 0.875%. All animals in the groups given 1.75, 3.5, or 7% TCP (with the same chemical composition as the NTP study above) in the diet exhibited piloerection, tremors, and diarrhoea, and were lethargic before death during the 14 day exposure. (Chapin et al. 1988). Dermal contact For mixed TCP isomers, the LD50-value has been reported to be > 7900 mg/kg in rabbits and 1500 mg/kg in cats. 4.1.4 Study on o-TCP Groups of three male cats received a single dose of 0, 100, 250, 500, 1000, 1500, or 2000 mg o-TCP/kg b.w. on a 15 cm2 clipped area of the neck. There was no wash off after the dosing. In order to reduce the acute effects of o-TCP dosing, cats treated with 1000 mg/kg and higher doses received 2 ml subcutaneous injections of atropine sulphate and PAM (pyridine-2-aldoxime-methylchloride) immediately after o-TCP administration and then one to three times daily until the disappearance of acute cholinergic effects of o-TCP. The cats were observed for up to 120 days with respect to body weight, development of OPIDN, and electromyographic changes. In the 2000 mg/kg b.w. group, two cats died after 6 and 7 days while the third one died on day 25 without recovering from the acute poisoning from o-TCP. Cats treated with 250 to 1000 mg o-TCP/kg b.w. lost weight after administration, and the weight loss was generally dose dependent. As time passed, these animals regained all of the lost weight and continued to gain weight until the end of the experiment. Cats given 500 or 1000 mg o-TCP/kg b.w. had significantly higher gains (157 and 141% of initial weight, respectively) than those of the untreated control group, which at termination was 124% of the cats initial weight. The functional disturbances of delayed neurotoxicity were always more severe in the hind legs than the fore limbs. The onset of these showed a dose dependent latency period being 17 days in the 1500 mg/kg b.w. group and 27 days in the 250 mg/kg b.w. group for leg weakness, and 31 to 36 days for mild ataxia of these two groups. In the 100 mg/kg b.w. group no signs of delayed neurotoxicity developed. Histopathologic examination of spinal cord showed changes in virtually all cats of the 250 to 2000 mg/kg b.w. groups. Histopathological examination of the peripheral nerves showed changes in cats of the 1000 to 2000 mg/kg b.w. groups. In the 100 mg/kg b.w. group no histopathologic signs of delayed neurotoxicity were observed. In electromyography, pathologic changes were observed in the 250 to 1500 mg/kg b.w. dose groups starting on day 37 to 50 with no apparent dose relationship. In the 100 mg/kg b.w. group no electromyographic change was observed. (Abou-Donia et al. 1986). 4.2 Long term toxicity Inhalation No data have been found. Oral administration As o-TCP and other isomers with at least one o-cresyl group are causing OPIDN, studies with o-TCP will be described before studies, where the content of o-TCP is low or not specified. Further as there is an OECD guideline for testing for OPDIN in 90 day studies using hens, the hen study will be described first. 4.2.1 o-TCP In 90-day studies in hens, functional (ataxia) and morphological neuropathological changes were found at daily oral dose levels of 5 to 20 mg o-TCP/kg b.w., but not at 2.5 mg/kg b.w./day and lower dosages. The degree of ataxia correlated well with the degree of neuropathological changes (types not specified). The functional changes were graded on a scale from 0 to 8, with the score 0 representing no ataxia and score 8 representing a situation, where the bird is unable to stand, have weak limb movements, and tail and leg reflexes are virtually non-existent. (Prentice et al. 1983 and Roberts et al. 1983). In order to test whether the absolute dose or the dosing regime (low daily (group A) versus high biweekly (group B) doses) were important for the neurotoxic action of o-TCP, a group of 40 male Long Evans rats received 116 mg o-TCP/kg b.w. by gavage 5 days/week for 24 weeks, another group of 70 rats received every two weeks a dose of 1160 mg o-TCP/kg b.w. for a period of 24 weeks, and finally a control group of 30 animals received 1 ml corn oil/kg. Group B animals received a prophylactic dose of 5 mg atropine sulphate 4 h after o-TCP dosing. Six to 10 animals from each group were killed for morphological study at 2, 6, 12, 18 and 24 weeks of exposure. Every six weeks, the rats were tested for plantar extension (placing reflex), when placed on a flat surface, general hindlimb co-ordination, when walking, and hindlimb splay. Over 60% group B animals survived the acute cholinergic (acetyl choline esterase inhibition) effects. Body weights increased by 54, 67, and 86% in group A, B and the control group, respectively. Functional abnormalities of group B animals were first noted at week 6 in the form of curled hind limb claws. After 12 weeks exposure, the same animals walked on their heels in stead of their toes and failed to maintain lateral extension of their hindlimbs, when lifted by the tail. After 18 weeks, these animals developed a peculiar rasping sound to their squeak, when excited by handling. After 24 weeks, over 50% of group B animals showed hindlimb crossing, when lifted by the tail. Functional abnormalities in group A animals rarely went beyond abnormalities in walking e.g. heel walking even after 24 weeks of exposure. Animals from both dosed groups showed visible loss of the thigh muscles after 24 weeks exposure. In morphological and histopathological studies, the neuronal damage was most pronounced in group B, with the ascending and descending tracts being damaged concurrently. At week 12, a study was carried out to see, which axons of the PNS were most profoundly affected. In both dosed groups the large diameter fibres, particularly tibial nerve branches, were preferentially damaged. (Veronesi 1984). This study shows that few dosings employing larger amounts of o-TCP given weeks apart produce more severe neurological toxicity, both behavioural and pathological, than smaller daily dosings, even though the same total dose was given to both groups. Groups of 10 male Fischer 344 rats received 0, 10, 50, or 100 mg o-TCP/ kg b.w./day orally in corn oil for 63 days. A positive control group of three chickens were given oral doses of 100 mg o-TCP/kg b.w./day for 18 days. The rats were tested in four sensorimotor tests after the last dose and after sacrifice, half of the rats were used for neuropathological studies and the other half for biochemical assays. In the 50 and 100 mg/kg b.w./day groups body weight gain was slightly reduced and signs of acute cholinergic toxicity (diarrhoea, salivation, and ocular discharge) were observed in these two dose groups only. At no time during the dosing period did rats show signs of delayed neurotoxicity. Chickens treated with o-TCP developed ataxia on day 7 and paralysis on day 15 of treatment. In rats, the peak intensity of movement, horizontal motor activity, and vertical motor activity were non-significantly dose-related different from control values. In all dosed rats, forelimb grip strength was significantly decreased (p<0.05). Brain acetylcholinesterase activity was significantly and dose-dependently decreased in the 50 and 100 mg/kg b.w./day groups with 20.1% of the control value being found in the 100 mg/kg b.w./day group. Brain NTE activity was also dose-dependently decreased in the 50 and 100 mg/kg b.w./day groups with 34.2% of the control value being found in the 100 mg/kg b.w./day group. Neuropathological changes revealed swollen axons without fragmentation or loss of myelin staining in the spinal cord of only one rat receiving 100 mg/kg b.w./day. (Somkuti et al. 1988). 4.2.2 Other TCP’s Groups of 10 male and 10 female F344/N rats received tricresyl phosphate in corn oil by gavage at doses of 0, 50, 100, 200, 400, or 800 mg/kg b.w./day for 13 weeks (NTP 1994) The tricresyl phosphate used in this and the subsequent NTP studies contained 21% m-TCP, 4% p-TCP and less than 0.1% o-TCP. Gas chromatographic analysis revealed two major peaks (24 and 30% of the total area) being isomers, for which the chemical composition was not revealed. Dicresyl phosphate esters comprised 17% of the mixture. All rats survived to the end of the study. Final mean body weights of male rats receiving 200, 400, and 800 mg/kg b.w./day were significantly lower than that of the controls. Cytoplasmic vacuolisation of the adrenal cortex occurred in all dosed groups and the severity increased with dose. Ovarian interstitial cell hypertrophy occurred in all dosed groups of females, and atrophy of the seminiferous tubules occurred in male rats that received 400 and 800 mg/kg b.w./day. There was a significant change in hindlimb grip strength in female rats that received 400 and 800 mg/kg b.w./day accompanied by a statistically lower body weight in 200, 400, and 800 mg/kg b.w./day females. Groups of 10 male and 10 female F344/N rats received a diet containing 0, 900, 1700, 3300, 6600, or 13000 ppm tricresyl phosphate for 13 weeks, equivalent to average daily doses of 55, 120, 220, 430, and 750 mg/kg b.w./day for males and 65, 120, 230, 430, and 770 mg/kg b.w./day for females (NTP 1994). All rats survived to the end of the study. Final mean body weights of males and females given 6600 and 13000 ppm and females given 3300 ppm TCP were significantly lower than those of controls. Feed consumption by male and female rats given 13000 ppm were significantly lower than that by controls during the first week of the study. In male rats at 13000 ppm a significant reduction in hindlimb grip strength was observed, possibly caused by the reduced body weight (120 g less than control). Cytoplasmic vacuolisation of the adrenal cortex occurred in all exposed groups of rats. Hypertrophy of ovarian interstitial cells and inflammation of the ovarian interstitium occurred in all dosed groups of females. Renal papillary oedema and renal papillary necrosis occurred in 13000 ppm males and females and in 6600 ppm females. Basophilic hypertrophy of the pituitary gland pars distalis and atrophy of the seminiferous tubules occurred in 6600 and 13000 ppm males. Groups of 95 male and 95 female F344/N rats received diets containing 0, 75, 150, or 300 ppm TCP for two years (NTP 1994). The average daily doses were 3, 6, and 13 mg/kg b.w. to males and 4, 7, 15 mg/kg b.w. to females. An additional group of 95 male and 95 female rats were given diets containing 600 ppm TCP for 22 weeks and then received only control feed. After 3, 9, and 15 months of exposure, up to 15 males and 15 females per group were evaluated for forelimb and hindlimb grip strength, then necropsied and evaluated for histopathologic lesions. Survival, feed consumption, and final mean body weights of all groups of rats given TCP were similar to those of the controls. Cytoplasmic vacuolisation of the adrenal cortex, which might be a general indication of stress, occurred in 300 ppm males and in 75, 150, and 300 ppm females at the 3 months interim evaluation. At 9 and 15 months, cytoplasmic vacuolisation occurred only in female rats, primarily in the 300 ppm group. Cytoplasmic vacuolisation of the adrenal cortex and ovarian interstitial hyperplasia occurred in female rats given 300 ppm TCP throughout the 2-year study and the incidence and severity increased at the end of the study. No dose-related neurotoxicity was observed. Groups of 10 male and 10 female B6C3F1 mice received TCP in corn oil by gavage at doses of 0, 50, 100, 200, 400, or 800 mg/kg b.w./day for 13 weeks (NTP 1994). All mice survived to the end of the study. Final mean body weight of male mice receiving 200 mg/kg b.w./day and of male and female mice receiving 400 and 800 mg/kg b.w./day were significantly lower than those of the controls. Cytoplasmic vacuolisation of the adrenal cortex occurred in all dosed groups and the severity increased with dose. Ovarian interstitial cell hypertrophy was present in all dosed female mice. Multifocal degeneration of the spinal cord occurred in males and females that received 100, 200, 400, and 800 mg/kg b.w./day, and multifocal degeneration of the sciatic nerve occurred in males that received 200, 400, and 800 mg/kg b.w./day and females that received 100, 200, 400, and 800 mg/kg b.w./day. Hind limb grip strengths of male mice that received 200, 400, and 800 mg/kg b.w./day were significantly lower than that of controls at the end of study. (NTP 1994). Groups of 10 male and 10 female B6C3F1 mice were fed diets containing 0, 250, 500, 1000, 2100, or 4200 ppm tricresyl phosphate for 13 weeks (NTP 1994). Dietary levels of 250, 500, 1000, 2100, and 4200 ppm TCP were equivalent to average daily doses of 45, 110, 180, 380, and 900 mg/kg b.w./day to males and 65, 130, 230, 530, and 1050 mg/kg b.w./day to females. All mice survived to the end of the study. Mean body weights of 4200 ppm males and of 2100 and 4200 ppm females were lower than those of the controls throughout the study. Feed consumption in females given 1000, 2100, or 4200 ppm TCP was lower than that by controls during week 12. Cytoplasmic vacuolisation of the adrenal cortex occurred in all dosed groups of males and females with the exception of 250 ppm males. Papillary hyperplasia of the gallbladder mucosa occurred in male mice given 500 ppm or more and in female mice given 1000 ppm or more. Axonal degeneration occurred in males and females given to 2100 and 4200 ppm and in females of the 1000 ppm group. Renal tubule degeneration occurred in all 4200 ppm male mice, and not in any other groups. The interpretation of the grip strength changes observed in the groups receiving 2100 or 4200 ppm TCP were confounded by the reduced body weight of these groups. There was a significant loss of both fore- and hindlimb grip strength in both sexes of the high dose group. (NTP 1994). Groups of 95 male and 95 female mice were fed diets containing 0, 60, 125, or 250 ppm TCP for two years (NTP 1994). The dietary levels were estimated to correspond to average daily doses of 7, 13, and 27 mg/kg b.w. to males and 8, 18, and 37 mg/kg b.w. to females. At 3, 9, and 15 months of chemical exposure, up to 15 males and 15 females per group were evaluated for forelimb and hindlimb grip strength, then necropsied and evaluated for histopathologic lesions. Survival, feed consumption, and final mean body weights of males and females given TCP were similar to those of the controls. Incidences of clear cell foci, fatty change, and ceroid pigmentation of the liver were significantly increased in male mice that received 125 or 250 ppm TCP. There were no statistically significant differences in fore- and hindlimb grip strength at the 3, 9, and 15 months examinations. Dermal contact 4.2.3 o-TCP Groups of at least three male cats received daily dermal doses of 0, 0.5, 1, 5, 10, or 100 mg o-TCP/kg b.w. for 90 days. There were no washing off of the substance. The surviving cats were further observed for a 30-day period. Cats that developed acute toxicity signs received daily 1-ml subcutaneous injections of atropine and PAM. In the 100 mg/kg/day group body weight decrease was observed steadily throughout the survival period of the cats which lasted to day 40, when the cats of this group were killed for humane reasons. In the 10 mg/kg/ day group body weight initially fell, and during the recovery phase there was an increase in body weight. In the 5 mg/kg b.w./day group body weight at the end of dosing was 135% of the initial body weight compared to 122% for the control group. The 1 mg/kg b.w./day and the 0.5 mg/kg b.w./day groups had a body weight of 116 and 110% of the initial body weight at the end of dosing. Signs of toxicity (cholinergic effects) developed in the 100, 10, and 5 mg/kg b.w./day groups 9, 25, and 35 days after onset of dosing, respectively. In the two last groups recovery from toxicity began 35 and 45 days after dosing, respectively. In the 1 and 0.5 mg/kg b.w./day groups no sign of acute toxicity was seen. Delayed neurotoxicity developed in the 100 mg/kg b.w./day group on day 23 with mild ataxia and developed further to paresis on day 33. In the 10 mg/kg b.w./day group leg weakness started on day 21 and developed further to paresis on day 40 in one cat. In the 5 mg/kg b.w./day group leg weakness started on day 52 and developed into mild ataxia on day 65. In the 1 mg/kg b.w./day group leg weakness started on day 74 and had recovered on day 92. No sign of delayed neurotoxicity was seen in the 0.5 mg/kg b.w./day group. In four cats of the 10 mg/kg b.w./day group electromyography was carried out at intervals. Electromyographic changes were seen in the cats from day 49-51 and recovery had taken place on day 68-72. Comparing the onset of clinical signs and electromyographic signs of delayed neurotoxicity, it appears that the clinical signs appear first. (Abou-Donia et al. 1986). 4.3 Reproductive and developmental effects 4.3.1 o-TCP Groups of three roosters received 0 or 100 mg o-TCP/kg b.w./day for 18 days, whereafter reproductive parameters were assessed. Two groups of roosters received 750 mg o-TCP/kg b.w. once and were sacrificed day 1 and day 18 after treatment, respectively. In the roosters sacrificed day 1 after dosing no clinical signs were observed. In the animals sacrificed day 18 developed typical signs of OPIDN at approximately day 12. In the animals given daily doses of 100 mg/kg b.w., testis weight and percent motile sperm were significantly decreased, whereas those animals that received a single dose of 750 mg/kg b.w. and were sacrificed on day 18 no adverse effects on these parameters were seen. In 5 of 10 roosters treated with 100 mg/kg b.w./day for 18 days, there was a significant disorganisation of the seminiferous epithelium, which affected 20-80% of the tubules per animal. The remaining 5 roosters of this group showed no consistent pattern of pathology. The two single dose o-TCP (1 and 18-day sacrifice) treatment groups showed no histopathological damage in the testis. (Somkuti et al. 1987). Groups of 10 to 18 pregnant Long-Evans rats received 0, 87.5, 175, 350 mg o-TCP/kg b.w. from day 6 to 18 of gestation. Standard OECD guideline teratology laboratory procedures were used in the assessment of the teratogenic potential of o-TCP. In the 350 mg/kg b.w. group 28% (5) of the animals died and one had totally resorbed foetuses. No maternal deaths or toxicity were observed in the 87.5 and 175 mg/kg b.w. groups. There were no significant differences noted among the dosed groups and the control group for preimplantation loss or resorption. Foetal weights for both sexes in the o-TCP dosed groups were significantly greater than in the control group. The results of this study indicate that o-TCP is not teratogenic in the Long-Evans rat. (Tocco et al. 1987). Groups of 12 male Long-Evans rats received 0, 100, or 200 mg TCP (less than 9% o-TCP)/kg b.w./day by gavage in 10.0 ml corn oil for 56 days prior to breeding and throughout a 10 day breeding period and groups of 24 female Long-Evans rats received 0, 200, or 400 mg TCP/kg/day for 14 days prior to breeding, and throughout breeding, gestation, and lactation. The males and females were mated - one male to two females - as follows: control to control, low dose to low dose, and high dose to high dose. During the mating period females were checked for sperm every morning until sperm cells were detected, whereafter the mated females were housed individually. No clinical signs or body weight depression were observed in any of the TCP dosed groups relative to controls. The observed rate of sperm positive females was similar in all female groups i.e. mating capacity was not affected. However, the percent of sperm positive females giving birth dropped from 95% in the control group to 45% in the low dose group and further to 5 % in the high dose group females i.e. fertility was affected by TCP dosing. Pups of the only litter in the high dose group died on lactation day 5 due to dehydration, and no milk was found in their stomachs. An evaluation of reproductive organ weights and sperm parameters was carried out on male rats at the end of dosing. Epididymis weight and sperm concentration in the high dose group were significantly decreased and in both the low and high dose groups motility, progressive movement and sperm morphology were adversely affected. Histopathologic changes were observed in the testes and epididymides of male rats and in the ovaries of female rats exposed to TCP. (Carlton et al. 1987). 4.3.2 Other TCP’s Male F344 rats and male B6C3F1 mice were treated with 0, 50, 100, or 200 mg/kg b.w. (mice), 0, 50, 100, 200, 400, or 800 mg/kg b.w. (rats) TCP containing <0.1% o-TCP by gavage in corn oil, or 0, 1700, 3300, or 6600 ppm (rats) or 0, 500, 1000, or 2100 ppm (mice) TCP in feed for 13 weeks. Sperm concentration, motility, and morphology were evaluated. The reproductive tract was examined for histopathologic lesions. Mice exposed by gavage exhibited a decrease in sperm concentration and an increase in
abnormal sperm morphology at 200 mg/kg b.w.; a decrease in sperm motility was found in all
TCP dosed groups. Tricresyl phosphate treatment significantly affected (p<0.01) all
male reproductive parameters in a dose-dependent manner in rats given TCP by gavage. In
animals given TCP in the feed, sperm parameters were adversely affected in the 2100 ppm
group mice and 6600 ppm group rats. Preliminary histopathologic findings indicate
multifocal testicular degeneration in rats but not mice by both dosing routes. At
comparable dose levels (mg/kg b.w.) corn oil gavage administration of TCP results in
greater male reproductive toxicity in rats and mice than dosing in the diet. In a continuous breeding protocol pairs of 20 F344 rats received 0.4 g TCP/kg b.w./day for 7 days prior to mating, through a 63 day breeding period, a 28-day post breeding interval and a crossover mating period. The TCP used is stated to be free of o-TCP and consisted of p- and m- isomers (62 wt%), cresyl-xylyl (18 wt%) and cresyl-ethylphenyl (18 wt%) phosphates. A group of 40 pairs served as control group. During the breeding period only 9 litters were delivered by the TCP dosed animals compared to 107 litters being delivered by the control group i.e. fertility was affected by TCP dosing. In the crossover mating, where control group males were mated to TCP dosed females and vice versa, it was shown that TCP caused infertility in TCP dosed males but not in the TCP dosed females. (Latendresse et al. 1994). In a continuous breeding protocol, groups of at least 20 pairs of CD1 mice received 0, 0.05, 0.1, or 0.2% TCP in the diet for 98 days, equivalent to 62.5, 124, and 250 mg/kg b.w./day. Fertility parameters and maternal body weight were used to assess the reproductive toxicity of TCP, containing 74.9% TCP with virtually no o-TCP (<0.1%). The remaining 25% were dicresyl phenyl and di- and tricresylxylyl phosphates. In the 0.1 and 0.2% dose groups reproductive toxicity was seen in the form of reduced numbers of pairs giving litters (for the 0.1% dose at the last litter and for the 0.2% dose group from the second litter and onwards). In the 0.2% dose group the numbers of live pups were decreased, and the numbers of dead pups were increased significantly. In the 0.1 and 0.2% dose groups the combined mean live pup weight was significantly decreased. In order to determine which sex was affected, control males were mated to 0.2% females, and 0.2% males were mated to control females. In a control group control males were mated to control females. No difference was observed in mating index, but the fertility index was reduced in both of the tested groups compared to the control pairs. Additionally the combined number of live pups per litter was significantly reduced in both test groups most in the combination control male mated to 0.2% female. This indicates that fertility of both sexes was affected by the 0.2% TCP dosing. In sperm parameters the 0.2% TCP dosed males had reduced percentage of motile sperm,
reduced sperm concentration and increased percentage of abnormal sperm. 4.4 Mutagenic and genotoxic effects Technical TCP (no further specifications given) was tested in the Ames test by two independent laboratories using Salmonella typhimurium strains TA100, TA1535, TA 1537, and TA98 with and without metabolic activation. TCP did not show mutagenic activity in the tests. (Haworth et al. 1983). TCP was tested for its ability to induce Unscheduled DNA Synthesis (UDS) in hepatocytes from male Fischer-344 rats after in vivo dosing by gavage and after in vitro exposure. All of the tests were negative. No information on doses in the in vivo study and concentration of TCP in the in vitro culture is provided in the abstract. (Mirsalis et al. 1983). NTP (1994) tested TCP for genotoxicity in three different in vitro genotoxicity tests with and without metabolic activation. In the Ames test Salmonella typhimurium strains TA98, TA100, TA1535, and TA1537 was used and TCP was tested in concentrations up to 10,000 m g/plate. In hamster ovary cell cultures sister chromatid exchange and chromosomal aberrations was studied in at least four different doses of TCP with 5,000 m g/ml as the highest concentration. None of the tests gave positive evidence for in vitro genotoxicity of TCP. 4.5 Carcinogenic effects Groups of 95 male and 95 female F344/N rats received diets containing 0, 75, 150, or 300 ppm TCP for two years (NTP 1994). The dietary levels were estimated to correspond to average daily doses of 3, 6, and 13 mg/kg b.w. to males and 4, 7, 15 mg/kg b.w. to females. No carcinogenic effects were observed in the study. Groups of 95 male and 95 female B6C3F1 mice were fed diets containing 0, 60, 125, or 250 ppm TCP for two years (NTP 1994). The dietary levels were estimated to correspond to average daily doses of 7, 13, and 27 mg/kg b.w. to males and 8, 18, and 37 mg/kg b.w. to females. No carcinogenic effects were observed. 5. Regulations, limit values
6. SummaryDescription Tricresyl phosphate (TCP) is the common name for mixtures of 10 different isomers, where the ortho-, meta- and para-cresyl groups can vary in number from zero to three. Industrial-grade tricresyl phosphate contains predominately the meta- and para-isomers and modern mixtures contain less than 1% of the ortho-isomer. Irrespective of the isomer composition, TCP’s used for industrial purposes are generally high boiling, inflammable liquids with a wide variety of uses. Environment TCP does not occur naturally in the environment but due to its many uses it can be found in air at concentrations up to 70 ng/m3, in water at levels of about 4 ng/l, and in soil at contaminated sites up to 4 mg/kg have been found. In sewage sludge up to 12 mg/kg dry matter have been detected. TCP has also been found in fish and shellfish in concentrations up to 40 ng/g. TCP released into water is readily adsorbed on to sediment particles, and little or none remains in solution. TCP is relatively immobile in soil as it adsorbs strongly to soil and is not expected to leach. Biodegradation is the dominant degradation process in aerobic waters and in soil. The degradation pathway most probably involves stepwise enzymatic hydrolysis to orthophosphate and phenolic moieties; the phenol would then be expected to undergo further degradation. Among the isomers of TCP, the ortho isomer degraded in river water slightly faster than the meta isomer and both isomers degraded faster than the para isomer. Toxicokinetics TCP’s are readily absorbed from the intestine of rats after oral dosing. Absorption after dermal application varies greatly depending on species with poor absorption in the dog, better absorption in man, and even better absorption in cats. After absorption it is distributed to the various organs with the highest concentrations in liver, blood, kidney and lung. o- and p-TCP are metabolised in essentially the same way to hydroxybenzoic acid. For o-TCP there is a "bypass", as a ring structure can be formed after the first cresyl moiety has been split off. This results in the formation of saligenin cyclic o-tolyl phosphate, a substance, which is at least five times more neurotoxic that o-TCP. No studies on the metabolism of m-TCP have been found. For the three isomers o-TCP, m-TCP-, and p-TCP up to 90-100% of an oral dose was excreted in urine and faeces within three days. At a dose of 2 mg/kg b.w. elimination was predominantly in urine, whereas at the dose of 200 mg/kg b.w. elimination was predominantly via faeces. Human toxicity Neurotoxicity and especially organophosphorous induced delayed neurotoxicity (OPIDN) is the critical effect of o-TCP in man. A single dose of 0.15 g o-TCP (2 mg/kg b.w.) is stated to produce toxic symptoms (type not specified) in man and severe neurological disturbance (OPIDN) developed after a single dose of 0.5-0.7 g o-TCP (7-10 mg o-TCP/kg b.w.). After short term ingestion of 5-10 mg TCP/kg b.w./day severe neurotoxic symptoms and/or OPIDN were seen. The TCP was stated not to contain o-TCP. After multiple ingestion of an estimated dose of 47.5-76 mg TCP containing 20% o-TCP/day, mild nausea and weakness were seen in 85 individuals and OPIDN in 91 persons. After long term ingestion of TCP often of unknown o-isomer content several reports of OPIDN have been described involving up to 50,000 people. Animal toxicity A summary of all short and long term animal studies is given in tables as an appendix. oral, short term TCP and the pure tricresyl esters have low acute lethal toxicity, however, large species differences exist, apparently with chicken/hens being the most sensitive species and rats and mice being the least sensitive species. After a single oral dose to hens, TCP’s not containing an o-cresyl ring are without capability of inducing OPIDN. If one o-cresyl ring is present, the substance is ten times more potent than o-TCP. Substances with two o-cresyl rings are 5 times more potent that o-TCP in inducing OPIDN. After a single oral dose of 100 mg o-TCP/kg b.w. cats, dogs, cows, sheep and chicken are susceptible to OPIDN-related paralysis (the critical effect of o-TCP in man), whereas rats and mice are less susceptible to the ataxia but susceptible to the pathological changes. A number of studies are available to assess the short term oral toxicity of TCP isomers. However, only one can be used to set a clear NOAEL. In that study, a NOAEL of 0.875% TPC in the diet applied to mice, as piloerection, tremors, and diarrhoea were observed at higher doses. oral, long term In 90 day studies using hens, ataxia and neuropathological changes were found at oral doses of 5 to 20 mg o-TCP/kg b.w./day. These are considered the critical effects of o-TCP. The NOAEL from these studies is 2.5 mg/kg b.w./day. In long term studies, the chronic toxicity of TCP-isomers with less than 0.1% o-TCP has been described. The best are the two-year studies, which also provide the lowest NOAELs. In the 2-year rat study, a NOAEL of 150 ppm in the diet equal to 7 mg/kg b.w./day was found based on the occurrence of ovarian interstitial hyperplasia in females at the higher doses. No signs of neurotoxicity were seen in this study. In the 2-year mouse study, a NOAEL of 60 ppm in the diet equal to 7 mg TCP/kg b.w./day can be set based on the occurrence of a significant increase in early signs of liver toxicity (clear cell foci, fatty change, and ceroid pigmentation) at the higher doses. No signs of neurotoxicity were seen in this study. Further good quality studies are the four 13 week studies in rats and mice, please refer to appendix. Reproductive and developmental effects A dose of 100 mg o-TPC/kg b.w./day caused pathological changes in rooster testes (disorganisation of the seminiferous epithelium) and affected the fertility of male rats in a one-generation study. In a teratogenicity study o-TCP did not show teratogenic effects in rats at doses up to 175 mg/kg b.w./day. In male rats and male mice given TCP by gavage for 13 weeks, the NOAEL for reproductive effects (semen parameters) was below 50 mg/kg b.w./day. In a mouse continuous breeding study a NOAEL for effects on fertility (reduced number of pups) was set at 62.5 mg/kg b.w./day. Mutagenic and genotoxic effects TCP has been tested in a number of in vitro and one in vivo genotoxicity test without showing genotoxic potential. Carcinogenicity TCP has been tested for carcinogenicity in two-year rat and mouse feeding studies without finding evidence for carcinogenicity. 7. Evaluation7. Evaluation As appears from the preceding sections, there are striking differences in the toxicity of different isomers of tricresyl phosphate, the differences being dependent on the occurrence of one or more or, as the second alternative, no o-cresyl moieties in the molecule. Therefore, in this and the following sections, the results involving studies with o-TCP and those with TCP containing less than 0.1% o-TCP will be dealt with separately. 7.1.1 o-TCP The toxicological data relating to the toxicity of o-TCP is mainly restricted to effects on the nervous system, as this organ for long has been considered the target organ for o-TCP toxicity. The studies include both human case stories and animal studies. The human data base consists of a number of studies, where poisoning and or development of organophosphorous induced delayed neurotoxicity (OPIDN) have occurred. However, only few of them have assessed the actual dose of o-TPC ingested. Neurotoxicity and especially organophosphorous induced delayed neurotoxicity (OPIDN) is the critical effect of o-TCP in man. A single dose of 0.15 g o-TCP (2 mg/kg b.w.) is stated to produce toxic symptoms (type not specified) in man and severe neurological disturbance (OPIDN) developed after a single dose of 0.5-0.7 g o-TCP (7-10 mg/kg b.w.). After repeated intake of TCP with a varying content of o-TCP, a number of cases have been described. In a recent, well described incident, mild nausea and weakness were seen in 85 individuals and OPIDN in 91 individuals who had ingested flour contaminated with TCP (containing 20% of o-TCP); the intake of TCP by the affected individuals has been estimated to 47.5-76 mg/day (equivalent to 0.16 to 0.25 mg o-TCP/kg b.w./ day assuming a body weight of 60 kg and 20% o-TCP in the TCP). In a variety of animal species, several studies have shown that single gavage dosing of 100 mg o-TPC/kg b.w. causes OPIDN. However, most of the animal studies are not suitable for setting a NOAEL for acute toxicity, as only a single dose was used. In the hen, which is considered to be the most relevant animal species for extrapolating the results to humans with respect to OPIDN, a NOAEL of 2.5 mg/kg b.w./day in 90-day oral studies applies; at the next dose level (5.0 mg/kg b.w./day), OPIDN developed. For estimation of a TDI for o-TCP, the LOAEL of 0.16 mg o-TCP/kg b.w./day for OPIDN in humans will be used. This LOAEL is below the dose levels causing neurotoxicity in humans and animals following acute exposure and below the NOAEL for OPIDN in the hen. This LOAEL is also clearly below the LOAEL for reproductive effects in roosters (100 mg/kg b.w./day) and is considered also to safeguard against reproductive effects. 7.1.2 TCP’s containing less than 0.1% o-TCP There is only one report on human poisoning after ingestion of TCP stated not to contain o-TCP. Here neurotoxic symptoms/OPIDN was observed in women after two weeks ingestion of 5-10 mg TCP/kg b.w./ day. However, in the short original description, no information on the content of mono- or di-o-cresyl phosphates is given. It is considered likely that the neurotoxic symptoms observed could be caused by these substances or by other substances capable of inducing OPIDN. Therefore, this study is not considered adequate for estimating a TDI for TCP containing less than 0.1% o-TCP. The most comprehensive studies on TCP toxicity have been carried out by NTP. These include a number of short and long term studies in mice and rats as well as a reproductive toxicity study using mice and genotoxicity tests. In the two year rat study, a NOAEL of 7 mg/kg b.w./day was observed with ovarian interstitial cell hyperplasia occurring at higher dose levels. Ovarian interstitial cell hyperplasia represents one of the first signs of reproductive toxicity and is considered the critical effect in female rats. In male rats, no critical effects were observed. No dose related neurotoxicity, including grip strength change, was observed in this study. In the two year mouse study, liver changes (clear cell foci, fatty change, and ceroid pigmentation) were observed in male mice with a NOAEL of 7 mg/kg b.w./day. The liver changes seen are generally considered to be the first signs of liver toxicity and are thus considered as being of toxicological significance, i.e. for male mice the critical effect is liver changes. In female mice, no critical effects were observed. No dose related neurotoxicity, including grip strength change, was observed in this study. From these studies, a NOAEL of 7 mg TCP/kg b.w./day applies for systemic toxicity. In the short term studies and also in 13-week studies, neurotoxic symptoms (most often reduced grip strength) were seen at high doses in some studies. These effects are likely to be caused by impurities of TCP containing one or two o-cresyl rings or the reduced grip strength could be caused by the reduced body weight seen at these doses. Whichever is the cause is not considered important to get clarified as clear systemic toxicity was observed in both rats and mice in two-year studies at dose levels of one to two orders of magnitude below those dose levels, where reduced grip strengths were observed. For estimation of a TDI for TCP containing less than 0.1% o-TCP, the NOAEL of 7 mg TPC/kg b.w./day from the two-year rat and mouse studies will be used as it safeguards against the neurotoxic effects seen in some studies at higher doses. 8. TDI, health based limit values8. TDI, health based limit values 8.1 TDI o-TCP The TDI of o-TCP is calculated based on the incident reported by Wang & Tao (1995) in which a LOAEL of 0.16 mg o-TCP/kg b.w./day was estimated for OPIDN in humans. = 0.0003 mg/kg b.w./day The safety factor SFI is set to 1 as human data are used. The SFII is set to 10 to protect the most sensitive individuals in the population. The SFIII is set to 50 as a LOAEL is used in stead of a NOAEL, because of the uncertainties in estimating a LOAEL from the data available, and because of the uncertainty on the content of mono-o-cresyl phosphates and di-o-cresyl phosphates (which are more potent than o-TCP in inducing OPIDN) in the contaminated flour. If the TDI is calculated based either on the NOAEL for o-TCP of 2.5 mg/kg b.w./day for functional (ataxia) and morphological neuropathological changes observed in the 90-day oral study in hens (Prentice et al. 1983, Roberts et al. 1983), or on the NOAEL for o-TCP of 0.5 mg/kg b.w./day for delayed neurotoxicity observed in the 90-day dermal study in cats (Abou-Donia et al. 1986), TDIs of 0.005 mg/kg b.w./day or 0.001 mg/kg b.w./day, respectively, are obtained taking a safety factor of 500 into account. Thus the data obtained in animal studies are in concordance with the available human data. TCP containing less than 0,1% o-TCP The TDI for TCP will be calculated based on the 2-year dietary studies in rats and mice (NTP 1994), where a NOAEL of 7 mg TCP/kg b.w./day was estimated for ovarian interstitial cell hyperplasia in the rat and for effects on the liver in the mouse. = 0.07 mg/kg b.w./day The safety factor SFI is set to 10 assuming that humans are more sensitive than animals. The SFII is set to 10 to protect the most sensitive individuals in the population. The SFIII is set to 1 as well conducted 2-year studies in two animal species form the basis for the estimated NOAEL. 8.2 Limit value in soil o-TCP Based on the TDI of 0.0003 mg/kg b.w./day and assuming a daily ingestion of 0.2 g soil for a child weighing 10 kg (wchild), a limit value is calculated: = 15 mg/kg soil This limit value will also take into account the possibility for a child of developing acute toxic effects following a single ingestion of up to 10 g soil. TCP containing less than 0,1% o-TCP Based on the TDI of 0.07 mg/kg b.w./per day (for a single commercial product of TCP) and assuming a daily ingestion of 0.2 g soil for a child weighing 10 kg (wchild), a limit value is calculated: The quality criteria for soil is going to cover all commercial products of TCP (containing less than 0.1% o-TCP) which consequently imply that different compositions of mixtures may occur and thus a possibility for mixtures containing more toxic isomers than the commercial product on which the TDI is based. Moreover, TCP are widespread in the environment having been detected in considerable amounts in sewage sludge. A further reduction factor (RF) is therefore set to 10. = 350 mg/kg soil 8.3 Limit value in drinking water o-TCP Based on the TDI of 0.3 µg/kg b.w./day and assuming a daily intake of 2 litres of water for a person weighing 70 kg (wadult), a limit is calculated: = 11 µg/l TCP containing less than 0,1% o-TCP Based on the TDI of 70 µg/kg b.w./day and assuming a daily intake of 2 litres of water for a person weighing 70 kg (wadult), a limit is calculated: The quality criteria for drinking water is going to cover all commercial products of TCP (containing less than 0.1% o-TCP) which consequently imply that different compositions of mixtures may occur and thus a possibility for mixtures containing more toxic isomers than the commercial product on which the TDI is based. Moreover, TCP are widespread in the environment having been detected in considerable amounts in sewage sludge. A further reduction factor (RF) is therefore set to 10. = 250 µg/l 9. Quality criteria9. C-value 9.1 Quality criteria in soil o-TCP A limit value of 15 mg/kg has been calculated for o-TCP based on children’s ingestion of soil. A quality criterion of 15 mg/kg soil is proposed. TCP containing less than 0.1% o-TCP A limit value of 350 mg/kg has been calculated for TCP containing less than 0.1% o-TCP based on children’s ingestion of soil. A quality criterion of 350 mg/kg soil is proposed. Quality criteria 15 mg/kg soil for o-TCP. 350 mg/kg soil for TCP containing less than 0.1% o-TCP. 9.2 Quality criteria in drinking water o-TCP A limit value of 11 mg/l has been calculated for o-TCP based on intake of drinking water. A quality criterion of 10 mg/l is proposed. TCP containing less than 0.1% o-TCP A limit value of 250 mg/l has been calculated for TCP containing less than 0.1% o-TCP based on intake of drinking water. A quality criterion of 250 mg/l is proposed. The limit value of 10 µg/l for AOX should also be observed. Quality criterion 10 mg/l for o-TCP. 250 µg/l calculated for TCP containing less than 0.1% o-TCP and 10 µg/l for AOX. 10. ReferencesAbou-Donia MB, Trofatter LP, Graham DG and Lapadula DM (1986). Electromyographic, neuropathologic, and functional correlates in the cat as the result of tri-o-cresyl phosphate delayed neurotoxicity. Toxicol Appl Pharmacol 83, 126-141. Amoore JE and Hautala E (1983). Odor as an aid to chemical safety: Odor thresholds compared with threshold limit values and volatilities for 214 industrial chemicals in air and water dilution. J Appl Toxicol 3, 272-290. At (1996). Grænseværdier for stoffer og materialer. Arbejdstilsynets At-anvisning Nr. 3.1.0.2, december 1996. Carlton BD, Basaran AH, Mezza LE and Smith MK (1987). Examination of the reproductive effects of tricresyl phosphate administered to Long-Evans rats. Toxicology 46, 321-328. Carlton BD, Irwin R, Hejtmancik M and Deskin R (1986). Reproductive toxicity of tricresyl phosphate in male rats and mice by two dosing routes. Toxicologist 6, 292. Chapin RE, George JD and Lamb JCIV (1988). Reproductive toxicity of tricresyl phosphate in a continuous breeding protocol in Swiss (CD-1) mice. Fundam Appl Toxicol 10, 344-354. EHC (1990). Tricresyl phosphate. Environmental Health Criteria 110. World Health Organization, International Programme on Chemical Safety, Geneva. Haworth S, Lawlor T, Mortelmans K, Speck WINGDINGS and Zieger E (1983). Salmonella mutagenicity test results for 250 chemicals. Environ Mutagen 5, 3-142. Henschler D (1959). Beziehungen zwischen chemischer Struktur und Lähmungswirkung von Triarylphosphaten. Naunyn-Schmiedeberg’s Arch exp Path u Pharmak 237, 459-472. Inoue N, Fujishiro K, Mori K and Mastuoka M (1988). Triorthocresyl phosphate poisoning - A review of human cases. J UOEH 10, 433-442. Latendresse JR, Brooks CL, Flemming CD and Capen CC (1994). Reproductive toxicity of bytulated triphenyl phosphate and tricresyl phosphate fluids in F344 rats. Fundam Appl Toxicol 22, 392-399. Merck Index (1996). An enclyclopedia of chemicals, drugs, and biologicals. 12th ed., Whitehouse Station, New Jersey, Merck & Co., Inc. Mirsalis J, Tyson K, Beck J, Loh E, Steinmetz K, Contreras C, Austere L, Martin S and Spalding J (1983). Induction of unscheduled DNA synthesis (UDS) in hepatocytes following in vitro and in vivo treatment. Environ Mut 5, 482. MM (1988). Bekendtgørelse om vandkvalitet og tilsyn med vandforsyningsanlæg. Miljøministeriets bekendtgørelse nr. 515 af 29. August 1988. MM (1997). The Statutory Order from the Ministry of the Environment no. 829 of November 6, 1997, on the List of Chemical Substances. MST (1990). Begrænsning af luftforurening fra virksomheder. Vejledning fra Miljøstyrelsen nr. 6 1990. MST (1996). Anvendelse af affaldsprodukter til jordbrugsformål. Bilagsdel til Miljøprojekt 328. Arbejdsrapport Nr. 47 1996. Nanda S and Tapaswi PK (1995). Biochemical, neuropathological and behavioural studies in hens induced by acute exposure of tri-ortho-cresyl phosphate. Intern J Neuroscience 82, 243-254. Norris P and Storrs FJ (1990). Allergic contact dermatitis to adhesive bandages. Dermatol Clin 8, 147-152. NTP (1994). Toxicology and carcinogenesis studies of tricresyl phosphate (CAS NO. 1330-78-5) in F344/n rats and B6C3F1 mice (gavage and feed studies). National Toxicology Program Technical Report Series no 433. OECD (1993). OECD Guidelines for the testing of chemicals, Volume 2. OECD, Paris. Patty’s Industrial Hygiene and Toxicology (1994). Fourth edition. Vol. II D. Chapter 32, section 11, p. 3063-3085 by Bisesi MS. Perry DF, Carter DE and Sipes IG (1983). Comparative excretion profiles of 3 tricresyl phosphates after oral administration to Fischer 344 rats. Toxicologist 3, 83. Prentice DE and Roberts NL (1983). Acute delayed neurotoxicity in hens dosed with tri-ortho-cresyl phosphate (TOCP): Correlation between clinical ataxia and neuropathological findings. Neurotoxicol 4, 271-276. Roberts NL, Fairley C and Phillips C (1983). Screening, acute delayed and subchronic neurotoxicity studies in the hen: Measurements and evaluations of clinical signs following administration of TOCP. Neurotoxicol 4, 263-270. Senanayaka N and Jeyaratnam J (1981). Toxic polyneuropathy due to gingili oil contaminated with tricresyl phosphate affecting adolescent girls in Sri Lanka. Lancet, January 10, 88-89. Somkuti SG, Lapadula DM, Chapin RE, Lamb JCIV and Abou-Donia MB (1987). Testicular toxicity following oral administration of tri-o-cresyl phosphate (TOCP) in roosters. Toxicology Lett 37, 279-290. Somkuti SG, Tilson HA, Brown HR, Campbell GA, Lapadula DM and Abou-Donia MB (1998). Lack of delayed neurotoxic effect after tri-o-cresyl phosphate treatment in male Fischer 344 rats: Biochemical, neurobehavioural, and neuropathological studies. Fundam Appl Toxicol 10, 199-205. Staehelin R (1941). Ueber Triorthokresylphosphatevergiftungen. Schweiz Med Wochenschr 16, 1-9. Tocco DR, Randall JL, York RG and Smith MK (1987). Evaluation of the teratogenic effects of tri-ortho-cresyl phosphate in the Long-Evans hooded rat. Fundam Appl Toxicol 8, 291-297. Tosi L, Righetti C, Adami L and Zanette G (1994). October 1942: a strang epidemic paralysis in Saval, Verona, Italy. Revision and diagnosis 50 years later of tri-ortho-cresyl phosphate poisoning. J Neurol Neurosurg Psychiatry 57, 810-813. Veronesi B (1984). A rodent model of organophosphorous-induced delayed neuropathy: Distribution of central (spinal cord) and peripheral nerve damage. Neuropathol Appl Neurobiol 10, 357-368. Wang D and Tao Y (1995). Toxic polyneuropathy due to flour contaminated with tricresyl phosphate in China. Clin Toxicol 33, 373-374. Wilson RD, Rowe LD, Lovering SL and Witzel DA (1982). Acute toxicity of tri-ortho-cresyl phosphate in sheep and swine. Am J Vet Res 43, 1954-1957. Appendix.Summary of short and long term toxicity of o-TCP.
Summary of short and long term toxicity data for TCP
April 1999 / final Evaluation of health hazards Benzoic acidand estimation of a limit value in air. Inger Thorup The Institute of Food Safety and Toxicology Danish Veterinary and Food Administration Denmark 1. General description 2. Toxicokinetics 3. Human toxicity 4. Toxicity, animal data 8. TDI, health based limit values 1. General description
1. General description 1.1 Identity Molecular formula: C7H6O2 Structural formula:
1.2 Physical / chemical properties
1.3 Production and use Benzoic acid is produced by air oxidation of toluene, hydrolysis of benzotrichloride, and decarboxylation of phthalic anhydride. (Merck Index 1996, HSDB 1998). The major use of benzoic acid is as preservative in foodstuffs, beverages industrial products, medicines and cosmetics. Further, uses have been found for benzoic acid in resin preparations, in production of plasticisers, in dyestuffs, in synthetic fibres, as a chemical intermediate, as corrosion inhibitor in paints and as plugging agent in oil well applications. In 1991 about 120000 tonnes of benzoic acid were produced in US and in 1988 about 233000 tonnes produced in Western Europe. (Kirk-Othmer 1985, BUA 1995, HSDB 1998). 1.4 Environmental occurrence Air Benzoic acid may be released into the environment as emissions. It is formed in combustion processes and found in automobile exhaust, refuse combustion and tobacco smoke. It will largely appear as aerosols. (HSDB 1998). Only very sparse data are available concerning atmospheric concentrations of benzoic acid. Concentrations at 1-26 ppt (10 ppt mean; corresponding to 0.05 µg/m3) has been measured in city air (HSDB 1998) and 0-0.30 ppm (approx. 0- 1.5 mg/m3) was found in an industrial environment (Halvorson 1984). In side-stream smoke from tobacco a maximal emission of 0.027 mg benzoic acid/cigarette was measured. Due to dilution of the smoke no benzoic acid could be detected in indoor air by currently available analysis (Patty 1993). Water More commonly benzoic acid is released in wastewater during its production and use in manufacturing of other compounds. Benzoic acid has in a few cases been detected in drinking water and ground water. (BUA 1995, HSDB 1998). Foodstuffs Benzoic acid occurs in nature in free and combined forms, especially in berries and fruit, which may contain up to 30 mg/kg. Benzoic acid and its sodium salt have over many years been extensively used as a preservative in concentrations up to 0.1% in foodstuffs. The intakes from natural sources are low in comparison with potential intakes from food additive uses. (SCF 1994, Merck 1996). 1.5 Environmental fate Air In the atmosphere benzoic acid will largely appears as aerosols, be subject to gravitational settling and be scavenged by rain (HSDB 1998). Water If released in wastewater benzoic acid seems readily biodegradable (half-life 0.2-3.6 days). Adsorption to sediment and volatilisation seems not to be significant. (IUCLID 1996, HSDB 1998). Soil If released on land, benzoic acid will percolate into the ground due to its low soil adsorption and biodegrade. It is rapidly metabolised by micro-organisms to benzylaspartic acid, benzoylglucoside, salicylic acid and its glucoside, and other unknown compounds (HSDB 1998). Bioaccumulation Bioconcentration in fish and algae does not seem important because of the low logP value (HSDB 1998). 1.6 Human exposure The general population is exposed to benzoic acid primarily through the ingestion of foods which contains benzoic acid naturally as well as food where benzoic acid is added as a preservative. Based on data regarding the amounts of benzoic acid and sodium benzoate produced in U.S. as food additives, a daily average exposure up to 4.4 mg benzoic acid /kg was calculated (IRIS 1997). In addition to the oral route, minor exposure may result from inhalation via auto exhaust, tobacco smoke and other combustion sources. Occupational exposure is primarily related to dermal contact or inhalation. The exposure may take place in pharmaceutical or chemical plants and the workers may be exposed to benzoic acid as dust or vapours. 2. Toxicokinetics2. Toxicokinetics 2.1 Absorption, distribution Inhalation No data after inhalation have been found. Oral, subcutaneous and intraperitoneal exposure Benzoic acid is rapidly and almost completely absorbed after oral administration in man and oral, intraperitoneal and subcutaneous administration in animals. (JECFA 1983, IUCLID 1996). Dermal contact After dermal application of 14C-labelled benzoic acid to humans in doses of 4, 40, and 2000 µg/cm2 a percutaneous absorption at 42.6%, 25.7% and 14.4%, respectively was measured. Studies in rhesus monkeys as well as humans showed that the greater topical doses of benzoic acid the less percutaneous absorption. Studies have shown that percutaneous absorption and absorption rate of benzoic acid is greater in human than in hairless dogs, whereas the absorption in guinea pigs is similar to that in humans (Maibach & Wester 1989 - quoted from IUCLID 1996). Transdermal absorption of benzoic acid has been studied in vitro in excised human skin and compared to absorption in living man. In equivalent time, 44.9% and 42.6%, respectively of the applied dose was absorbed (Franz 1975 - quoted from IUCLID 1996). Percutaneous absorption of 14C-labelled benzoic acid in aged people (> 65 years) was lower than in young people (18-40 years). The cumulative dose absorbed within seven days was 19.5% and 36.2%, respectively (Roskos et al. 1989 - quoted from IUCLID 1996). 2.2 Elimination Metabolism Benzoic acid is conjugated in the liver and two metabolites of have been detected: hippuric acid and benzoyl-glucuronic acid. Endogenous formation of benzoic acid takes place in the organism where phenylalanine and tyrosine are precursors. Experiments with labelled phenylalanine have shown that about 1-2% is metabolised to benzoic acid. This leads to excretion of a few tens of mg benzoic acid/kg bw/day in human urine. (JECFA 1983). Excretion Benzoic acid is nearly completely excreted in urine in animals as well as man. Less than 1% appears in the faeces. The excretion occur mostly in the form of hippuric acid (the glycine conjugate). A smaller part is excreted as benzoyl-glucuronic acid (the glucuronyl conjugate) and free benzoic acid. In man and most experimental animals the liver is the main site of conjugation. Depletion of glycine is the limiting rate of excretion of hippuric acid. Administered benzoic acid (p.o., i.p., s.c.) is excreted within 24 hours. Tissue accumulation does not occur. (JECFA 1983). In swine dermal exposure lead to urinary excretion of about 20% and faecal excretion of about 3% of the administered dose within six days (Carver & Riviere 1989 - quoted from IUCLID 1996). 2.3 Toxicological mechanisms Depletion of glycine, due to the formation of the main metabolite glycine conjugate (hippuric acid), may explain some of the adverse effects seen in experimental animals and man. Experiments in rats have shown that addition of glycine to the diet reduced the toxicity of benzoic acid (Griffith 1929 and Kowalewski 1960 - quoted from JECFA 1983). Although glycine is not an essential amino acid it appears that in growing animals or in animals subjected to injury there is a narrow margin between the metabolic demand of glycine and the rate at which glycine is formed or made available in the body. Further, shortage of glycine in the body can lead to disturbances in the acid-base equilibrium due to increased concentrations of benzoic acid in the organism. High doses of benzoic acid per se may also lead to disturbances in the body acid-base balance. (JECFA 1983). 3. Human toxicity3. Human toxicity Sodium benzoate is rapidly hydrolysed to benzoic acid in the body and there seems no reason to believe that sodium benzoate differs toxicologically from benzoic acid. Therefore, in the following some data based on investigations of sodium benzoate are included where relevant to describe the toxicological profile of benzoic acid. Conversion factor: 100 mg sodium benzoate is equivalent to 84.7 mg benzoic acid. 3.1 Short term toxicity Inhalation No data have been found. Oral intake A number of very early studies have shown gastro-intestinal disturbances (dyspepsia, nausea and vomiting) after single oral administration of 1-5 g sodium benzoate/benzoic acid to humans, corresponding to approx. 14-70 mg /kg bw expressed as benzoic acid. (Meissner & Shepard 1866 - quoted from JECFA 1983 and Goodman & Gilman 1966). A more recent study has confirmed these early stated values: Nausea, bloating, and epigastric discomfort have been observed in humans after oral administration of 4 g sodium benzoate corresponding to doses from 39-71 mg/kg bw (expressed as benzoic acid) (Jackson et al. 1987). Humans given 25 mg/kg bw for 20 days suffered from irritation (probably gastro-intestinal), discomfort, weakness and malaise (Wiley & Bigelow 1908 - quoted from SCF 1994). The effects were described as reversible and maybe caused by a disturbance of acid-base balance rather than local tissue damage. Some individuals are hypersensitive to benzoic acid. In such sensitive persons oral doses corresponding to 10-250 mg benzoic acid (less than 4 mg/kg bw) have been shown to induce eczema (Vieluf et al. 1990 and BIBRA 1989 - quoted from IUCLID 1996). Anaphylaxis was induced in a young female after ingestion of a meal containing sodium benzoate as preservative. A subsequent provocation test with oral administration of 160 mg sodium benzoate (corresponding to 2.5 mg benzoic acid/kg bw) induced localised urticaria (Michils 1991). It is known that persons who suffer from asthma, rhinitis, or urticaria may undergo exacerbation of symptoms following ingestion of foods or beverages containing benzoic acid or benzoates (JECFA 1983). A trained worker suffered from allergic reactions of increasing intensity while being constantly exposed to benzoic acid during work (no exposure levels are given). After oral exposure to 425 mg sodium benzoate (approx. 5 mg benzoic acid/kg bw) he developed anaphylaxis and showed milder reactions later when eating food containing benzoic acid (Pevny 1981). Dermal contact Benzoic acid may cause non-immunologic immediate contact reactions (erythema, local urticaria) within 30-45 min. of skin application. A number of skin tests with different doses in various vehicles have been carried out. (Forsbeck & Skog 1977, Rademaker & Forsyte 1989, Lewis et al. 1989, Ylipieti & Lahti 1989, Kligman 1990 and Larmi et al.,1989 - quoted from IUCLID 1996). 3.2 Long term toxicity Inhalation Benzoic acids in tailor chalk has been connected with occurence of rhinitis, shortness of breath and "burning eyes" of students and teachers from a fashion school. They were exposed to sublimated Benzoic acid as well as benzoic acid dust (Aberer 1992). The authors mentioned that after notification of the formula of this special chalk they could ascribe the problems to benzoic acid since only this chemical had been added for special purposes to the normal chalk. Duration and level of exposure was not mentioned. The exposure was intermittent and the symptoms disappeared when the persons were not longer exposed. Oral intake Very early studies have shown that doses up to 14 mg benzoic acid/kg bw/day for 88-92 days were without visible effect (Gerlach 1909- quoted from SCF 1994) and dietary administration of 0,3-0,4 g benzoic acid (4,3-5,7 mg/kg bw/day) for 62 days had no effect on blood picture, urine composition, nitrogen balance and well being (Chittenden et al. Quoted from JECFA 1983). Dermal contact Three workers of a pharmaceutical plant exposed to airborne sodium benzoate developed transient non-immunological contact urticaria related to skin contamination with sodium benzoate. No information on the concentration of the sodium benzoate in the air was given (Nethercott 1991). Eye irritation "Burning eyes" has been connected with exposure to benzoic acid in tailors chalk (Aberer 1992, see above). 3.3 Reproductive / developmental effects No data have been found. 3.4 Genotoxic effects No data have been found. 3.5 Carcinogenic effects No data have been found. 4. Toxicity, animal data4. Toxicity, animal data As for humans, the animal toxicity data include some information based sodium benzoate where relevant to describe the toxicological profile of benzoic acid. Conversion factor: 100 mg sodium benzoate is equivalent to 87.4 mg benzoic acid. A lot of animal exposure data exists on benzoic acid/sodium benzoate, however the majority of the studies mentioned below are of an earlier date and do not meet the requirement of today’s quality guidelines. For most of the studies the dose levels are given as percentage benzoic acid/sodium benzoate in the feed and therefore conversion from % in feed to mg/kg bw/day, which is needed for setting an ADI/TDI, has been carried out. In some cases this has been done according to a method used by IUCLID 1996, in other cases the following, widely used, calculation has been applied: Young rats consume 100 g feed/kg bw/day, old rats consume 100 g feed/kg bw/day and mice consume 150 g feed/kg bw/day. 4.1 Short term toxicity Inhalation An acute one hour inhalation toxicity study in rats using benzoic acid, stated to be in the vapour phase, has shown that LC50 is > 0.026 mg/l. The one hour exposure lead to generalised inactivity and lachrymation, but no mortality (Bio-Fax data sheet 1973 - quoted from IUCLID 1996). In a well performed four weeks study Sprague-Dawley rats (10 animals/ sex/group) were exposed by inhalation to benzoic acid (generated as a dust with an equivalent aerodynamic diameter of 4.7 µm). The animals were exposed to 0, 0.025, 0.25 or 1.2 mg benzoic acid/l for 6 hours/day, 5 days/week. All high- and mid-dose animals exhibited clinical signs of upper respiratory tract irritation (red material around the nares). Two animals (one/sex) in the high-dose group died. The cause of death was not determined. Statistically significant decrease in body weight gain, absolute and relative weights of liver, kidneys and trachea/lungs in high-dose animals were seen. Treatment related decrease in the number of blood platelets was seen in both sexes. The histopathological examination revealed compound-related lesions in the lungs: an increase in the intensity and extent of interstitial inflammatory cell infiltrate and an increase in the incidence and intensity of interstitial fibrosis. The lung lesions were seen in all dosed groups, and with respect to the cell infiltration the effect seems dose related. The author stated that the results of the study indicate that levels as low as 0.025 mg/l benzoic acid for four weeks produce toxic effects in the lungs (EPA/OTS 1992). Oral administration LD50-values Acute oral toxicity studies in mice, rats, rabbits and dogs have resulted in LD50 values within the range of 1700-2500 mg benzoic acid/kg bw. By i.p. and i.v. dosing slightly lower doses LD50 values were obtained. (Opdyke 1979, JECFA 1983, IUCLID 1996). Short-term studies The data are summarised in Table I. rats Groups of five male and five female Sherman rats were fed sodium benzoate for 30 days at levels ranging from 16-1090 mg/kg bw/day (corresponding to 14-923 mg benzoic acid/kg bw/day). A control group was included. No clinically adverse effects or histopathological changes in organs were detected. (Smyte & Carpenter 1948 - quoted from JECFA 1983). Twenty eight young rats (strain unknown) were given 0 or 5% sodium benzoate in the diet for three weeks. The animals consumed 39 g feed/ day which roughly corresponds to 3800 mg/kg bw/day expressed as benzoic acid. Nineteen animals died within two weeks and all others died within the following week. The food consumption was significantly reduced, the animals suffered from diarrhoea and crusted blood was seen around the nares. At autopsy intestinal bleeding was observed. (Kieckebusch & Lang 1960). Groups of three male and three female Sherman rats were fed 0, 2 and 5% sodium benzoate in the diet for 28 days (corresponding roughly to 1700 and 4250 mg/kg bw/day, expressed as benzoic acid). All animals on the 5% level died within the first two weeks showing hyperexcitability, urinary incontinence and convulsions. At the 2% level a decrease in body weight (only statistically significant for males) compared to that of controls was seen. No necropsy or histopathology data were reported. (Fanelli & Halliday 1963). Groups of 5-10 male Wistar rats were given 0 or 3% benzoic acid in the diet for five days (approx. 2250 mg/kg bw/day). After 4-5 days the animals showed disorders of the nervous system (excitation, ataxia and convulsions). After 3-5 days brain damage was demonstrated histologically. Another similar study with 15 male Wistar rats showed that the brain lesions were still present 19-30 days after cessation of dosing. (Kreis et al. 1967 - quoted from IUCLID 1996). Groups of 10 male and 10 female F344 rats were fed sodium benzoate in the diet at levels of 0, 0.5, 1, 2, 4 and 8% for six weeks (approx. 320-5120 mg /kg bw/day expressed as benzoic acid). All rats on the 8% level and 19 rats on the 4% level died within four weeks. The author reported that animals treated with sodium benzoate showed "hypersensitivity" (probably hyper reactivity) as an acute effect. No other symptoms were seen. A statistically significant reduction in body weight was noted in the two highest dose groups compared to control group. By necropsy atrophy of the spleen and lymphnodes was observed in rats fed 4 and 8% sodium benzoate. (Sodemoto & Enomoto 1980). Six male and six female F344 rats were fed 0, 1.81, 2.09 or 2.40% sodium benzoate for 10 days (approx. 1150, 1330 or 1525 mg/kg bw/day, expressed as benzoic acid). One male in the 2.40% group, who showed hypersensitivity followed by convulsions, died. The mean body weight of both sexes in the highest dose group was statistically significantly reduced compared with the controls. The relative liver and kidney weights were statistically significantly increased at the 2.40% level in both sexes and in the liver in males at the 2.09% level. Histopathological examination were carried out on liver and kidneys from the 2.40% level. Only the liver from high dosed males showed signs of toxic effect. Eosinophilic foci in the periportal area and enlargement of the hepatocytes with glassy cytoplasm was seen. Statistically significant changes in some clinical chemistry parameters (e.g. albumin) compared to controls were seen at the 2.09 and 2.40% level. (Fujitani 1993). Benzoic acid was given orally to 4 groups of 25 CDBR rats at doses of 0, 30, 160 or 450 mg/kg bw/day on days 7 to 16 of gestation. Four deaths occurred in the high dose group. Further, at this dose level reduction in body weight , reduced food intake and increased liver weight were seen. Haemorrhages in the gastric mucosa was seen in the rats which died. (EPA/OTS 1992 - abstract quoted from Toxline 1990-1993). mice Albino Swiss mice, 4 mice per sex per group, were given 0, 0.5, 1, 2, 4, or 8% sodium benzoate in drinking water (approx. 0, 1000, 2000, 4000, 8000 or 16000 mg/kg bw/day, expressed as benzoic acid) for 35 days. Based upon survival rate, body weight and histological changes (not specified) the 2% dose level were found suitable for a life span study (but not stated as a NOAEL). All animals at the 8% level and 3/4 in both sexes died at the 4% level. (Toth 1984). Five male and four to five female B6C3F1 mice were fed 0, 2.08, 2.5 or 3% sodium benzoate for 10 days (approx. 2550, 3180 or 3810 mg/kg bw/day, expressed as benzoic acid). All of the animals in the 3% group showed "hypersensitivity" (probably hyper reactivity) as an acute effect. and 1/5 of males and 2/5 of females showed convulsions. Two females died. Absolute and relative liver weights (both sexes) and relative kidney weights (females) were dose-dependently increased (statistically significant from controls at 3% the level). Histopathological examination were carried out on liver and kidneys from the 3% level. Only the liver from high dosed males showed signs of toxic effect. Enlarged hepatocytes with eosinophilic cytoplasm, occasionally single cell necrosis and vacuolation of hepatocytes was seen. Statistically significant changes in some clinical chemistry parameters (e.g. cholinesterase) compared to controls were seen at the 2.5 and 3% level. (Fujitani 1993). cats Cats have a higher sensitivity to benzoic acid compared with other species (including humans) which may be due to its lack of ability to form benzoyl glucuronic acid (JECFA 1983). The application of 0 or 0.5% benzoic acid in the diet for 3-4 days to four male cats (approximately 300-420 mg/kg bw/day) resulted in convulsions, aggression, and hyperaesthesia. Two cats died. A histological examination revealed toxic effects in the liver (swollen hepatocytes with cell infiltration) and kidneys (swollen tubular cells). (Bedford & Clarke 1972 - quoted from IUCLID 1996). Four male cats were given 0, 100 or 200 mg benzoic acid/kg bw/day in the diet for 15 days. In another study four male cats were fed a diet containing 0 or 0.25% benzoic acid (approximately 130-160 mg/kg bw/day) for 23 days. No effects were observed. (Bedford & Clarke 1972 - quoted from IUCLID 1996). An outbreak of poisoning in 18 cats following ingestion of meat containing 2.39% benzoic acid have been reported. The effects were convulsions, nervousness, excitability, and loss of balance and vision. Seventeen cats died or were killed. Damage to the intestinal mucosa and liver was seen at autopsy. (Bedford & Clarke 1971 - quoted from JECFA 1983). Table I. Results of short-term oral toxicity studies with benzoic acid or sodium benzoate.
* The test compound is given as benzoic acid or sodium benzoate. See the text for further specification. The bracket gives the value converted into benzoic acid. bb: benzoic acid. n.a.d: no abnormalities detected. n.d.r: no data reported. Dermal contact Acute dermal toxicity studies in rabbits have shown that the LD50 value of benzoic acid exceeds both 5000 mg (Opdyke 1979) and 10000 mg/kg bw (Bio-Fax data sheet 1973 - quoted from IUCLID 1996). No information concerning mortality is available for the first mentioned study. No deaths were observed among rabbits given > 10000 mg/kg bw. skin irritation From a skin irritation study on rabbits, performed according to relevant guideline, it was concluded that benzoic acid is minimally irritating to the skin (RRC NOTOX B.V. 1988 - quoted from IUCLID 1996). In a similar study design sodium benzoate did not cause skin irritation (RRC NOTOX B.V. 1988 - quoted from IUCLID 1996). skin sensitisation Different types of sensitisation tests have been carried out on guinea pigs and mice. No sensitising effect was seen. (Gad et al. 1986 and Gerberick et al. 1992 - quoted from IUCLID 1996). Eye irritation Eye irritation studies in rabbits performed according to relevant guidelines revealed slightly (benzoic acid) to severely (sodium benzoate) irritating properties (Suberg 1986 and RRC NOTOX B.V. 1988 - quoted from IUCLID 1996). 4.2 Long term toxicity Inhalation No data have been found. Oral administration The data are summarised in Table II. rats A 90-days feeding study was carried out on groups of 8-10 Sherman rats. The animals were given sodium benzoate at levels of 0, 1, 2, 4 or 8% in the diet (approx. 540-5160 mg /kg bw/day expressed as benzoic acid). In the 8% group 4/8 died. The four survivors showed reduced weight gain compared to that of the controls. Increased kidney and liver weights together with pathological lesions (not specified) were seen in the 8% group. At the lower levels no adverse effects were observed. (Deuel et al. 1954 - quoted from JECFA 1983 and IUCLID 1996). In a chronic feeding experiment continuing over 4 generations, three groups of 20 male and 20 female rats (strain unknown) were fed 0, 0.5 and 1% benzoic acid (1% benzoic acid mentioned as equal to 150 mg/rat/ day, roughly corresponding to 600 mg/kg bw/day). The third generation was subjected to histopathological investigation after 16 weeks on test and organ weight was recorded for brain, liver, heart, spleen, kidney and testes. 1% benzoic acid was tolerated without adverse effect on growth, food utilisation and duration of life. No differences with respect to organ weights and histopathological findings (organ not mentioned, but probably the same as recorded for organ weights) were seen between dosed and control animals. The authors stated that 1% benzoic acid in the diet is near the upper limit of tolerability. (Kieckebusch & Lang 1960). Twenty male and 30 female rats were fed a diet containing 1.5% benzoic acid (approx. 1125 mg/kg bw/day) for 18 months. 13 males and 12 females served as controls. Reduced food intake, growth retardation and increased mortality (15/50 vs. 3/25 in the control) was seen in the dosed animals. No pathological data was recorded. (Marquardt 1960 - quoted from JECFA 1983 and IUCLID 1996). mice 50 mice per sex were given 0 or 80 mg benzoic acid /kg/day by gavage for 12 weeks. Reduced weight gain without reduced food intake was observed. No post mortem data recorded. (Shtenberg & Ignat’ev 1970 - quoted from JECFA 1983 and IUCLID 1996). dogs Seventeen fox terriers were fed 0.1 - < 7 g benzoic acid or sodium benzoate (not otherwise specified) once daily during 250 days. Seven g/day (approx. 1000 mg/kg bw/day) induced ataxia, convulsions and death. Below this level no adverse effects were seen. (Rost et al. 1933 - quoted from JECFA 1983 and IUCLID 1996). Table II. Results of long-term oral toxicity studies with benzoic acid or sodium benzoate.
* The test compound is given as benzoic acid or sodium benzoate. See the text for further specification. The bracket gives the values converted into benzoic acid. bb: benzoic acid. n.a.d: no abnormalities detected. n.d.r: no data reported. Dermal contact No data have been found. 4.3 Reproductive / developmental effects In a four generation feeding experiment three groups of 20 male and 20 female rats (strain unknown) were fed 0, 0.5 and 1% benzoic acid (1% benzoic acid mentioned as equal to 150 mg/rat/day, roughly corresponding to 600 mg/kg bw/day). No effects on fertility and litter size neither at birth or weaning, were observed. (Kieckebusch & Lang 1960). Rats were given 0, 1, 2, 4, or 8% sodium benzoate (approx. 590-4720 mg/kg bw/day expressed as benzoic acid) in the diet during the whole gestation period (20 days). 4 and 8% sodium benzoate caused maternal toxicity, 100% perinatal death and many abnormalities of organs (eye, brain and kidneys) and skeletal system were found in foetuses from these groups. No statistically significant differences in organs and skeletal abnormalities were detected between the 1 and 2% levels and the controls. Maternal as well as teratogenic NOAEL was 2% corresponding to 1180 mg/kg bw/day. (Onodera et al. 1978 - quoted from IUCLID 1996). Benzoic acid was given orally to 4 groups of 25 CDBR rats at doses of 0, 30, 160 or 450 mg/kg bw/day on days 7 to 16 of gestation. Four deaths occurred in the high dose group. At this dose level reduction in body weight , reduced food intake and increased liver weight were seen. Reduction in body weight was also observed at 160 mg/kg/day. No compound-related effects on reproductive parameters (not specified) were observed. Foetal weight gain was significantly decreased at the two highest dose levels. Other effects reported in the high dose group were significant increases in malformations (not specified), foetal development variations (not specified), and variations due to retarded development (not specified). The maternal and foetal NOAEL reported in this study was 30 mg/kg/day. (EPA/OTS 1992 - abstract quoted from Toxline 1990-1993). In rats (up to 150 mg/kg bw/day on days 6 to 15 of gestation), mice (up to 150 mg/kg bw/day on days 6 to 15 of gestation), rabbits (up to 210 mg/kg bw/day on days 6 to 18 of gestation) and hamsters (up to 250 mg/kg bw/day on days 6 to 10 of gestation) sodium benzoate (value expressed as benzoic acid) was given orally. No effect on nidation or foetal and maternal survival was seen and the number of abnormalities of soft and skeletal tissues did not differ from controls. No maternal toxicity was reported. (Food and Drug Research Laboratories, Inc. - quoted from IUCLID 1996). It has been shown that benzoic acid penetrate the placental barrier readily after s.c. administration the maternal rats (Maickel & Snodgrass 1973 - quoted from IUCLID 1996). In an uterotrophic assay performed on mice and rats it was found that benzoic acid does not possess oestrogenic properties. This finding was supported by an in vitro test. (Ashby 1997). 4.4 Genotoxic effects In vitro tests mutagenicity Benzoic acid and sodium benzoate was not found mutagenic when tested in Ames test and E. coli reversion assay with and without metabolic activation (Ishidate et al. 1984, Zieger et al. 1988 and Prival et al. 1991 - all three quoted from IUCLID 1996). Ambiguous results was observed when sodium benzoate was tested in a B. subtilis recombination assay with and without metabolic activation (Ishizaki & Ueno 1989 - quoted from IUCLID 1996). chromosomal aberration Benzoic acid was tested in two different Chinese hamster cell lines without metabolic activation; negative or ambiguous results were obtained (Ishidate et al. 1984 - quoted from IUCLID 1996). Sodium benzoate was tested in two different Chinese hamster cell lines without metabolic activation; positive results were seen in both cell lines (Ishidate et al. 1977,1984 and Abe & Sasaki 1977 - quoted from IUCLID 1996). effects on DNA Sodium benzoate was tested for Sister Chromatide Exchange in Vicia faba root tip cells, hamster cell line and human lymphocytes without metabolic activation; a positive result in Vicia faba root tip cells and human lymphocytes was observed (but only one dose level tested). Ambiguous result was seen in the hamster study. (Abe & Sasaki 1977 and Xing & Zhang 1990 - quoted from IUCLID 1996). Benzoic acid was tested in human lymphocytes without metabolic activation; a negative result was obtained. (Tohda et al. 1980 and Jansson 1988 - quoted from IUCLID 1996). In vivo tests chromosomal aberration Rats were given 50-5000 mg sodium benzoate/kg orally for one to five days. No detectable significant aberrations of the bone marrow chromosomes. Negative results were also obtained in a dominant lethal assay using the same study design (Litton Bionetics Inc. 1974 - quoted from IUCLID 1996). 4.5 Carcinogenic effects rats Rats were fed 0, 1 or 2% sodium benzoate in the diet (approx. 500-1000 mg/kg/day) for 18-24 months. No differences in mortality were seen. No significant differences between groups with respect to number of tumour-bearing animals and time to occurrence of tumours. (Sodemoto & Enomoto 1980). mice Sodium benzoate was administered as a 2% solution in drinking water (approx. 4000 mg/kg/day) for the life span to 50 mice per sex. The control group consisted of 100 animals per sex. The author reported that no carcinogenic effect observed and the treatment has no effect on survival (Toth 1984). However, a higher incidence and earlier onset of mammary gland tumours was seen in dosed females (8%) compared to controls (2%). No statistical analysis were performed. The study is only reported as a short communication. 5. Regulations, limit values
6. SummaryDescription Benzoic acid appears at room temperature as colourless crystals or a white powder with no or a faint, pleasant odour. It is only slightly soluble in water, but soluble in alcohol and chloroform. The major use of benzoic acid is as preservative in food and industrial products. Environment Benzoic acid may be released into the environment as emissions and largely appears as aerosols. It is formed in combustion processes and found in automobile exhaust, refuse combustion and has been measured in tobacco smoke. Only very sparse data are available concerning atmospheric concentrations of benzoic acid. Concentrations at 1-26 ppt (in city air) and 0-0.30 ppm (industrial environment) have been measured. Benzoic acid is released into wastewater during its production and use in manufacturing of other compounds. In addition, benzoic acid occurs naturally in foodstuffs up to 30 mg/kg. Benzoic acid seems readily biodegradable in water and soil and does not accumulate. Human exposure The general population is mainly exposed to benzoic acid through the ingestion of foods which contains benzoic acid naturally or as a preservative. In addition to the oral route, exposure may result from inhalation via auto exhaust, tobacco smoke and other combustion sources. Toxicokinetics Benzoic acid is rapidly and almost completely absorbed after oral administration in man and animals. After dermal application to humans the absorption is about 40%. Within 24 hours benzoic acid is nearly completely excreted in the urine, mainly as hippuric acid. Endogenous formation of benzoic acid in the organism is taken place. No accumulation in the body. Human toxicity Sublimated benzoic acid as well as benzoic acid dust in the air has been connected with occurrence of rhinitis, shortness of breath and eye irritation. No data on duration and exposure was given. Benzoic acid has been reported to cause gastro-intestinal disorders after single oral or short term administration of 1-5 g, corresponding to approx. 14-70 mg /kg bw. However, doses below 14 mg /kg bw/day for 88-92 days were without visible effect. The effects seen in humans after ingestion of benzoic acid and sodium benzoate may be related to a disturbance in acid-base equilibrium (systemic effect) or to irritation of the surface epithelium (local effect). Besides the gastro-intestinal effect, benzoic acid is able to induce hypersensitivity reactions. Some persons who suffer from asthma, rhinitis, or urticaria undergo exacerbation of symptoms following ingestion of benzoic acid. Oral doses corresponding to less than 4 mg benzoic acid/kg bw has been shown able to cause skin reactions in sensitive persons. Anaphylaxis has been induced in sensitive persons after ingestion of a meal containing sodium benzoate as preservative and a following provocation test with oral administration of 160 mg sodium benzoate (corresponding to 2.5 mg benzoic acid/kg bw) induced localised urticaria. Dermal exposure to benzoic acid may induce transient non-immunological contact reactions (erythema, local urticaria). Animal toxicity A lot of animal exposure data exists on benzoic acid/sodium benzoate, however the majority of the studies mentioned below are of an earlier date and do not meet the requirement of today’s quality standards. An acute one hour inhalation toxicity study in rats using benzoic acid, stated to be in the vapour phase, has shown that the LC50value is > 26 mg/m3. In a four-week study rats were exposed by inhalation to benzoic acid dust aerosol. Doses of 25, 250 or 1200 mg/m3 for 6 hours/day, 5 days/week caused upper respiratory tract irritation in all mid- and high-dose animals. Deaths were seen at 1200 mg/m3. Statistically significant decrease in body weight gain and some organ weights were found in high-dose animals. The histopathological examination revealed compound-related lesions in the lungs: an increase in the intensity and extent of interstitial inflammatory cell infiltrate and an increase in the incidence and intensity of interstitial fibrosis. The lung lesions were seen in all dosed groups, and with respect to the cell infiltration the effect seems dose related. The author stated that the results of the study indicate that levels as low as 25 mg/m3 benzoic acid for four weeks produce toxic effects in the lungs. The acute oral toxicity of benzoic acid is rather low. LD50-values have been reported within the range of 1700-2500 mg benzoic acid/kg bw. The results of the short and long term toxicity studies in rats, mice, cats and dogs are given in table I and II. From the table it is seen that administration of benzoic acid or sodium benzoate at levels higher than about 1% in the diet for rodents (˜ 600 mg benzoic acid /kg/day), 0.5% for cats (˜ 400 mg/kg/day) and ˜ 1000 mg/kg/day for dogs resulted in adverse effects (increased mortality, changes in body and organ weight, damage to liver and kidney, neurological disorders). The LD50-value for dermal exposure is low (>10.000 mg/kg bw). Benzoic acid is found minimally irritating to the skin and slightly irritating to the eyes. Reproductive and developmental effects No human data have been found. In a four generation feeding experiment in rats benzoic acid in doses up to 1% benzoic acid (roughly corresponding to 600 mg/kg bw/day) did not lead to any effect on fertility and litter size. Up to 2% sodium benzoate in the diet during the whole gestation period (approx. 1180 mg benzoic acid/kg bw/day) was without adverse effect. At higher levels maternal toxicity, perinatal death and foetal abnormalities of organs and the skeletal system were found. Maternal toxicity and severe adverse effects on foetal development has been reported when benzoic acid was given to rats in doses of 450 mg/kg bw/day in day 7-16 of gestation, however, details were not specified (abstract only). Studies in mice, rabbits and hamsters did not show any signs of benzoic acid induced maternal or foetal toxicity. However, only dose levels up to 150-250 mg/kg/day were tested. Genotoxicity Benzoic acid/sodium benzoate was found without mutagenic effect in in vitro mutagenicity tests. However, with respect to sodium benzoate -but not benzoic acid- positive or ambiguous results were obtained in some of the in vitro chromosomal aberration tests and tests for effects on DNA. Negative results were seen in the in vivo chromosomal aberration tests. Carcinogenicity Two studies are available. According to the authors, no carcinogenic effect was observed. However, the studies do not meet the requirements of today’s guidelines and are inadequate for evaluation. 7. EvaluationStudies have shown that benzoic acid may induce adverse effects in humans as well as animals. However, most of the studies are of an earlier date and do not meet the requirement of today’s quality guidelines. Only a few studies reflect the effect of benzoic acid after inhalation. The sparse human data have shown that exposure to airborne benzoic acid may lead to reversible irritation of the respiratory system and eyes or skin reactions. However, none of these studies provide any exposure levels upon which a NOAEL/LOAEL can be set. A four-week well performed inhalation study in rats indicate that even at the lowest level tested (25 mg benzoic acid/m3) histopathological changes were observed. The treatment related effects were confined to the respiratory system and most likely of local nature. Oral studies in animals have shown that the acute toxicity of benzoic acid is rather low (LD50: 1700-2500 mg/kg bw). However, in humans sensitive to benzoic acid an acute systemic effect (anaphylaxis) has been observed after ingestion of a meal containing sodium benzoate as a preservative. Gastro-intestinal disorders in humans have been reported after single oral or short term administration of 1-5 g, corresponding to approx. 14-70 mg /kg bw. Doses below 14 mg/kg bw/day for 88-92 days were without visible effect. Most of the oral animal studies are inadequate and do not - or at best only very sparsely - include information about gross pathology and histopathology. Further, the mean benzoic acid consumption has been calculated according to different methods implying that the NOAEL/ LOAEL values are subject to a rather high degree of uncertainty. It is not possible to draw attention to one particular study for establishing a NOAEL. However, most of the data points toward the same effect: a systemic toxic effect which clinically appears as disorders of the nervous system or death at high dose levels and lesions to the liver and kidneys at lower levels. Administration of sodium benzoate or benzoic acid to mice and rats at levels below 1 % in the diet (approx. 600 mg/kg bw) seems without toxic effect. Levels above this may result in severe systemic disorders. The reprotoxicity studies are insufficient to draw any conclusions from, but exposure to levels of benzoic acid not leading to maternal toxicity probably do not cause developmental/foetal toxicity. Benzoic acid is not mutagenic and not genotoxic in vivo. The two carcinogenicity studies available do not meet the requirements of today’s guidelines and are inadequate for evaluation. Based upon the available data from human as well as animal studies, the critical effects after inhalation of benzoic acid is considered to be an irritative, most likely reversible effect on the respiratory system and eyes and skin reactions. For estimating a limit value in air, the LOAEL at 25 mg benzoic acid/m3 from the rat study is considered most relevant. Although the critical effect of benzoic acid is considered to be irritation to various tissue surfaces, it cannot be excluded that inhaled benzoic acid may cause an acute systemic effect (anaphylaxis) in humans sensitive to benzoic acid. 8. Limit value in airFor the estimation of a limit value in air, a LOAEL of 25 mg benzoic acid/m3 from the four-week rat study is considered.
= 0.025 mg/m3 The safety factor SFI is set to 10 as animal data are used. The SFII is set to 10 to protect the most sensitive individuals in the population. The SFIII is set to 10 since a LOAEL is used in stead of a NOAEL and because the calculation is based upon a short-term four-week study. 9. C-valueA limit value of 0.025 mg/m3 has been calculated. For substances having acute or subchronic effects, but for which activity over a certain period of time is necessary before the harmful effect occurs, the C-value is set at the limit value. A C-value of 0.02 mg/m3 and placing in Main Group 2 is proposed (MST 1992). C-value 0.02 mg/m3, Main Group 2. 10. ReferencesAberer VW, Kager B, Ziegler V and Horak F (1992). Schnupfen durch Schneiderkreide. Dermatosem 40 (6), 231-234. Ashby J, Lefevre PA, Odum J, Tinwell H, Kennedy SJ, Beresford N and Sumpter JP (1997). Failure to Confirm Oestrogenic Activity for Benzoic acid and Clofibrate: Implications for Lists of Endocrine-Disrupting Agents. Regul Toxicol Pharmacol 26, 96-101. BUA (1995). Beratergremium für umweltrelevante Altstoffe (BUA) der Geschellschaft Deutscher Chemiker. BUA-Stoffberichte 145-147 (Kurzberichte): Benzoesäure,Natruimbenzoat. Wissenschatlige Verlagsgesellschaft 1995. EPA/OTS(1992). Doc #88-920001538. Initial submission: INM-5268. Developmental toxicity study in rats dosed by gavage on days 7-16 of gestation (final report) with attachments and cover letter dated 032792. Fanelli GM and Halliday SL (1963). Relative toxicity of chlortetracycline and sodium benzoate after oral administration to rats. Arch Int Pharmacodyn 144 (1-2), 120-125. FAO/WHO (1962). Specifications for identity and purity of food additives. Vol. I Antimicrobial preservatives and antioxidants. Benzoic acid. Fujitani T (1993). Short-term effect of sodium benzoate in F344 rats and B6C3F1 mice. Toxicol Lett 69, 171-179. Goodman LS and Gilman A (1966). In: The pharmacological basis of therapeutics, 3rd. ed. The Macmillan Company, New York, 1030. Halvorson DO (1984). Summary Report. Determination of Benzoic Acid. Air AM Ind Assoc J 45, 727-730. HSDB (through 1998). Benzoic acid. In: Hazardous Substances Databank. IRIS (1997). Benzoic acid. In: Integrated Risk Information System. Database quest, last revised: 03-01-97. US-EPA. Jackson AA, Badaloo AV, Forrester T, Hibbert JM and Persaud C (1987). Urinary excretion of 5-oxoproline (pyroglutamic aciduria) as an index of glycine insufficiency in normal man. Br J Nutr 58, 207-214. JECFA 1983. WHO Food Additive Series No. 18. International Programme on Chemical Safety. Toxicological Evaluation of Certain Food Additives and Food Contaminants, 28-40. Kieckebusch VW and Lang K (1960). Dir Verträglichkeit der Benzoesäure im chronischen Fütterungsversuch. Arzmeim-Forsch 10, 1001-1003. Kingsett’s Chemical Encyclopaedia (1966). A digest of chemistry and its industrial applications. Hey, HD. (Ed). Ninth edition. Bailliére, Tindall and Cassell. London. Kirk-Othmer (1985). Benzoic acid. In: Concise Encyclopaedia of Chemical technology. John Wiley & Sons. A Wiley-Interscience Publication, 162-63. Merck Index (1996). Benzoic acid. In: an Encyclopedia of Chemicals, Drugs, and Biologicals. 12th ed., Rahway, New Jersey, Merck & Co., Inc., 1124. Michils A, Vandermoten G, Duchateau J and Yernault JC (1991). Anaphylaxis with sodium benzoate. The Lancet 337, 1424-1425. MST (1992). Industrial Air Pollution Control Guidelines. Vejledning fra Miljøstyrelsen nr. 9 1992. Opdyke DLJ (1979). Special Issue V. Monographs on Fragrance Raw Materials. Benzoic acid. Fd Cosmet Toxicol 17, 715-717. Nethercott JR, Murray J, Lawrence MJ, Roy AM and Gibson BL (1991). Airborne Contact Urticaria Due to Sodium Benzoate in a Pharmaceutical Manufacturing Plant J Occup Med 26, 734-736. Patty“s Industrial Hygiene and Toxicology (1993). Fourth Edition. Vol. II, Part A. Toxicology. Clayton GD, Clayton FE (Eds.). John Wiley and Sons, Inc., 111-112. Pevny VI, Rauscher E, Lechner W and Metz J (1991). Exzessive Allergie gegen Benzosäure mit anaphylaktischem Schock nach Expositionstest. Dermatosem, 29, 123-130. SCF (1994). Scientific Committee for Food. CS/CONS/30. Opinion on Benzoic acid and its salts. Sodemoto Y and Enomoto M (1980). Report of carcinogenesis bioassay of sodium benzoate in rats: Absence of carcinogenicity of sodium benzoate in rats. J Environ Path Toxicol, 4, 87-95. Toth B (1984). Short communications. Lack of Tumorigenicity of Sodium Benzoate in Mice. Fundam Appl Toxicol 4, 494-496.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||