Ethylbenzene

Endpoint summary

Additional information

Ethylbenzene is absorbed from the lungs, gastrointestinal tract and through the skin. Inhalation studies in humans demonstrate that up to 64 % of the inhaled ethylbenzene vapour can be retained in the lungs (Bardodej and Bardodejova, 1970), whereas for animals, slightly lower retention rates (44 %) have been reported (Chin et al., 1980b). Therefore, for risk assessment purposes, inhalation absorption percentages of 65 % for humans and 45 % for animals should be taken into consideration. It is recognized that the inhalation absorption rate in humans is based on an old study with limited reporting. But this absorption rate is supported by data for styrene, a substance with very similar physicochemical properties. Engström et al. (1978) determined the uptake of styrene after inhalation exposure to 210 mg/m³ (corresponding to 50 ppm ) in volunteers under different physical activity (0 – 150 Watt) to be 63%.

 

No studies are available regarding the absorption of ethylbenzene after oral exposure of humans. Studies in animals indicate that ethylbenzene is quickly absorbed via this route. Up to 92 % of ethylbenzene metabolites could be detected in the urine of animals within 24 hrs after dosing (El Masry et al., 1956). In rats 48 h after oral application 82.4% of the dose were recovered as metabolites in urine and 1.5% in feces, respectively (Climie et al., 1983). Therefore, as worst case, 100 % oral absorption in animals and humans is taken for risk characterisation.

 

Studies in humans dermally exposed to liquid ethylbenzene demonstrate rapid absorption through the skin, whereas no indications of absorption could be seen after exposure of humans to ethylbenzene vapors. Up to 45 % of ethylbenzene was absorbed in human volunteers, who had immersed one hand for up to 2 hrs in aqueous solutions of ethylbenzene (Dutkiewicz and Tyras, 1967). In predictions based on a dermal PBPK model, human dermal absorption rates were between 34 and 37 % (Shatkin and Brown, 1991). The modeling was based on conditions similar to those as used by Dutkiewicz and Tyras (1967), i.e.the skin being immersed in an aqueous solution of ethylbenzene for one hour. These findings would lead to a worst case assumption of 50 % dermal absorption. But these data cannot be used for dermal risk assessment and derivation of a dermal DNEL for two reasons: resorption from an aqueous solution certainly is different than that of undiluted ethylbenzene as encountered at ethylbenzene workplaces and an exposure duration of 1 or 2 hours without allowing for evaporation does not reflect actual exposure situations.

 

Experimental studies on dermal penetration should reflect actual exposure conditions. For volatile solvents evaporation after skin contact must be taken into account. Thus, it is stated in the OECD guidance document for the conduct of skin absorption studies (OECD 2004a) that in vivo as well as in vitro investigations should reflect the human exposure scenario. The OECD guidelines for testing skin absorption either in vitro (OECD 2004b) or in vivo (OECD 2004c) describe the design of a typical device for investigating volatile chemicals. After application the material is allowed to evaporate and is trapped by an activated charcoal filter for quantification. Radiolabelled material should be used for in vivo dermal penetration and the amount absorbed should be quantified by measuring radioactivity in metabolism cages (excreta, expired air, carcass). The rat is the most commonly used species, but hairless strains and species having skin absorption rates more similar to those of humans, can also be used. The comparability of skin absorption rates of humans and other species, especially hairhess strain, is substantiated by OECD (2004c) by the following references: ECETOC, 1993; EPA, 1992; EPA, 1998; Bronaugh et al., 1990; Feldman and Maibach, 1970.

 

Susten et al. (1990) determined the in vivo percutaneous absorption of ethylbenzene in hairless mice with an experimental design corresponding to OECD (2004c) with only minor deviations. Details of the procedure are described by Susten et al. (1986): A “skin depot” for application of the test material consists of a stainless-steal casing glued to the skin using cyanoacrylate adhesive with a treatment area of 0.8 cm², a charcoal basket on top of the casing allowing for trapping of evaporated solvent, and a teflon cap through which the test material is placed onto the skin by an application syringe. This approach allows evaporation of volatile substances and trapping within the charcoal basket for quantitative determination.

 

Susten et al. (1990) applied 5 µl of [14C] ethylbenzene onto the skin within the skin depot and the radioactivity remaining in the syringe was determined to calculate the amount actually applied. Directly after dosing the animals were placed into metabolism cages and sacrificed at the end of a 4 h treatment period. Radioactivity was determined in the skin-depot (charcoal trapping) for measuring of evaporated test substance, in a wipe of the treated skin site for unabsorbed chemical, in the skin treatment site for bound chemical, in the rest of the carcass, in feces, urine, cage wash, and in expired breath samples collected during 0-15, 15-30, 30-60, 60-120, 120-180, and 180-240 min after placing the animals into the metabolism cages. 3.61 % of the actually applied ethylbenzene was absorbed through the skin and 86.7 % of the nominal dose was recovered from the skin depot (evaporation into charcoal). Skin contact time during which absorption could occur (taking into account evaporation of ethylbenzene from the skin surface) was estimated to be 5 min with an absorption rate of 37 µg/cm²/min.

 

This approach followed closely the OECD guideline for in vivo skin absorption (OECD, 2004c): The application device allowed for evaporation of ethylbenzene and quantitative recovery by the charcoal filter, absorption was determined by measuring radioactivity in treated skin, carcass, excreta and expired air, adequate recovery of 100 ± 10 %, and use of hairless mice that are assumed to have absorption rates similar to those of humans according to OECD. The only major difference is the exposure time; OECD asks for 6 or 24 h while in this experiment 4 h were used. However, this difference is negligible since the skin contact time was estimated to be only 5 min and afterwards the material had evaporated into the charcoal. The use of hairless mice was justified by referring to a comparison of Bronaugh et al. (1982) of the percutaneous absorption of various chemicals through the skin of humans, miniature pigs, monkeys, guinea pigs, rabbits, rats, hairless mice, and normal mice. For hairless mice the ratio of permeability relative to humans was between 0.5 and 2.0 for 8 of 9 chemicals and for 1 chemical it was 4.9. On the basis of such a comparison the next best species was the pig with 8 of 13 ratios in the range of 0.5 and 2. This corresponds to the statement of OECD (2004c) that hairless species have skin absorption rates more similar to those of humans in comparison to the rat that is the most commonly used species.

 

Susten et al. (1990) noted the large differences in absorption rates when comparing their own data (37 µg/cm²/min) with those of Dutkiewicz and Tyras (1967) measuring the disappearance of ethylbenzene applied under a watch glass to the forearm of humans over 10-15 min (366-550 µg/cm²/min). The major criticisms were:

- the methodology was not well described

- evaporative losses during application and recovery of unabsorbed material may have been a source of error

- there were no independent measures of absorption (e.g. analyses of blood or exhaled breath concentrations or urinary excretion of metabolites)

- furthermore, the absorption rates for toluene obtained by the method of Dutkiewicz and Tyras (1967) were not consistent (much higher) with absorption rates derived from studies of other authors with toluene (references given by Susten et al., 1990).

 

In addition:

A comparison of the rat skin in vitro data of Tsuruta (1982) with the human in vivo data of Dutkiewicz and Tyras (1967) for undiluted ethylbenzene casts further doubt on the findings of the latter authors. Tsuruta (1982) used excised abdominal rat skin placed in a diffusion cell. Ethylbenzene was placed onto the skin and the diffusion cell was sealed with a glass stopper without allowance for evaporation. According to Dutkiewicz and Tyras (1967) the absorption rate for undiluted ethylbenzene was 28 mg/ cm²/h, corresponding to 470 µg/ cm²/min (most probably still within lag period before steady state is achieved). Tsuruta (1982) reported an absorption rate of 0.993 nmoles/cm²/min, corresponding to 0.1 µg/ cm²/min. The difference of a factor of nearly 5000 can neither be explained by the experimental setup (in vivo vs. in vitro), nor by the different species under investigation since animal skin is generally more permeable and therefore would rather overestimate human percutaneous absorption (OECD, 2004a).

 

In the UK Risk Assessment Report (2008) for styrene, a chemical with similar physical chemical properties, a dermal penetration of 2 % was taken for risk characterization (page 80/81): This was based on an in vitro skin penetration experiment with human skin according to OECD (2004b) allowing for evaporation of the volatile chemical and absorption by a charcoal filter. By this method the absorbed dose of styrene was 1.25 % being in the same range as the absorption determined by Susten et al. (1990) for ethylbenzene.

 

In summary, for practical use conditions a dermal absorption rate of 3.61 % should be the starting point for the risk assessment with 4 % as a worst case assumption.

Distribution

 

Ethylbenzene is rapidly distributed through the body. After inhalation exposure of animals, highest amounts of radioactivity in tissues after 6 hr exposure to 230 ppm ethylbenzene were found in the carcass, liver and gastrointestinal tract and lower amounts detected in the adipose tissue (Chin et al., 1980b). From humans exposed to a mixture of industrial xylene containing 40.4 % ethylbenzene it was derived, that the retention of ethylbenzene in adipose tissue was approximately 2 % of the total uptake (Engström and Bjurström, 1978). There was, however, no evidence of ethylbenzene accumulation in fat-rich tissues.

Metabolism and Excretion

Species differences have been shown concerning the metabolism of ethylbenzene. In humans exposed via inhalation, the major metabolites are mandelic acid (approximately 70 % of the absorbed dose) and phenylglyoxylic acid (approximately 59 % of the absorbed dose), which are excreted in the urine. Both metabolites are formed after side chain oxidation. In rats exposed by inhalation or orally, the major metabolites were identified as hippuric acid and benzoic acid (both metabolites amount to 38 % of the metabolites), 1-phenylethanol (25 % of the metabolites), mandelic acid (15 – 23 % of the metabolites) and phenylglyoxyloc acid (10 % of the metabolites). In rabbits, the most relevant metabolite is hippuric acid. Rabbits excrete higher levels of glucuronidated metabolites compared to rats and humans. Low levels of the glutathione-derived metabolites 1-phenylethylmercapturic acid, 2-ethylphenylmercapturic acid, 3-ethylphenylmercapturic acid and 4-ethylphenylmercapturic acid have been detected in rat urine after inhalation exposure to ethylbenzene vapour (Cossec et al., 2010). Human and animal data demonstrate, that ring oxidation (and formation of metabolites resulting from ring oxidation) represents a minor pathway compared to side chain oxidation.

In animals, ethylbenzene kinetics was saturable at concentrations above 500 ppm in mice (Charest-Tardif et al., 2006) and concentrations of 200 ppm in rats (Tardif et al., 1996).

Ethylbenzene is rapidly metabolized and metabolites are eliminated rapidly from the body, primarily as urinary metabolites. Excretion is almost complete within 24 hrs after exposure with only about 0.2 % of absorbed dose remaining in the body within 42 hrs after inhalation exposure. In man, highest urinary concentrations of mandelic acid and phenylglyoxylic acid occurred 7 hrs after exposure, the biological half-life of both metabolites is 4 – 7 hrs (Hagemann and Angerer, 1979). Comparable results were obtained for other metabolites, such as the ring oxidation product 2-ethylphenol and the phenolic metabolites m- and p-hydroxyacetophenone: within a period up to 8 hrs after exposure, more than half the quantity measured in 24 hr-urine could be determined (Engström et al., 1984).

After inhalation exposure, exhalation is also an important pathway of excretion, whereas fecal excretion plays a minor role (Chin et al., 1980b).

Saghir et al. (2009) conducted a comprehensive study to determine species differences in the metabolism of ethylbenzene by mouse, rat, and human lung and liver microsomes. The side chain oxidation to 1-phenylethanol and acetophenone and ring oxidation pathways to 2-ethylphenol, 4-ethylpenol, 2,5-ethylquinone, and 3,4-ethylquinone were measured. Reactive metabolites were monitored by glutathione trapping (2,5- and 3,4-dihydroxyethlybenzene-GSH). None of the metabolites were detected with human lung microsomes. Conversion to 1-phenylethanol ranged from 1% by rat lung to 58% by mouse lung microsomes. More 1-phenylethanol was found in mouse lung than in mouse liver microsomes while formation was similar by rat lung and liver microsomes. Metabolism to 1-phenylethanol was in the order mouse>rat>human. Formation of acetopheneone was about one order of magnitude lower than that of 1-phenylethanol. In a subsequent investigation (Saghir et al. 2010), the formation of ethylbenzene-derived reactive metabolites by microsomal fractions isolated from liver and lung was found to be greater in preparations from mouse relative to those from rat or human. This difference was particularly marked for lung fractions, where an approximate 15-fold difference in the extent of covalent binding was apparent for mice and rats, with an approximate 50-fold difference when mice and humans were compared. These results indicate species and tissue dependent differences in the formation of reactive metabolites, with the greater activity of mouse lung tissue consistent with findings of lung tumours in this species after long-term inhalation exposure. Conversion to ring hydroxylated metabolites was much lower. 2,5-dihydroxyethylbenzene-GSH was formed to the highest extent by mouse lung microsomes and to a higher extent by mouse liver than by rat or human liver microsomes. After incubation of 2- and 4-ethylphenol the conversion to ethylcatechol and ethylhydroquinone followed the order mouse>human>rat for liver microsomes and mouse>rat>human for lung microsomes. Ring oxidized metabolites accounted for only a relative small fraction of the overall ethylbenzene metabolism but it was selectively elevated in mouse lung microsomes.