Registration Dossier

Toxicological information

Direct observations: clinical cases, poisoning incidents and other

Currently viewing:

Administrative data

direct observations: clinical cases, poisoning incidents and other
Type of information:
experimental study
Adequacy of study:
supporting study
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
other: Non-guideline study published in peer reviewed literature, adequate for assessment. Klimisch rating 2 “reliable with restrictions

Data source

Reference Type:
Report date:

Materials and methods

Study type:
study with volunteers
Endpoint addressed:
basic toxicokinetics
Test guideline
no guideline followed
Principles of method if other than guideline:
To establish relationships between ethylene (ET) and ET oxide (EO) exposure scenarios in mice rats and humans, five physiologically based toxicokinetic (PBT) models have been described (Fennell and Brown, 2001; Hattis, 1987; Krishnan et al., 1992; Smith, 1988) and Csanády et al., 2000). Of the five, only Csanády et al., 2000 examined ET and EO. ET is found ubiquitously in the environment (74% in air from natural sources, 26% from burning biomass and incomplete combustion of fossil fuels). ET is an important industrial chemical, is a plant hormone, is produced endogenously in mammals and is metabolised to EO by mammals. EO is also an important industrial chemical. ET and EO are gaseous at room temperature and EO is mutagenic in prokaryotic and eukaryotic cells, is genotoxic and clastogenic in mice and rats and probably humans. However, exposure to ET is neither mutagenic, genotoxic or clastogenic in rodents. ET at high concentrations (≥20,000 ppm) causes acute hepatoxicity in rats that have been pre-treated with non-specific enzyme inducers e.g. Arochlor 1254. EO levels in mice or rats following single inhalation exposures up to 10,000 ppm for 6 hours resulted in blood EO concentrations of <100 ppm at 6 h, regardless of exposure dose. In volunteer humans, metabolically produced EO levels in blood were directly proportional to ET exposure up to the level tested (50 ppm) (Filser et al., 2013). Of the five PBT models published, the model of Csanády et al., 2000 most accurately represented ET and EO exposure. However missing data (EO concentrations in blood, inhalation kinetics of ET and EO in mice, ET metabolism kinetics and cytochrome P450 (CYP) inhibition by ET), meant that the model performed poorly when predicting EO exhaled by rats following high exposure (> 1000 ppm) to ET. The data gaps identified have been addressed in the literature since the publication of the Csanády et al., 2000 model.
This paper aims to build upon the Csanády et al., 2000 PBT model by
1) extending the model to include the novel information generated by incorporating additional compartments with metabolic competence, include CYP2E1 mediated metabolism of ET and ET inhibition of CYP activity, include microsomal epoxide hydroxylase (EH)- and glutathione S transferase (GST)- as well as non-enzymatic transformation of ET to EO, thus enabling a more realistic and accurate description of the disposition of ET and EO exposure in rats, mice and humans,
2) to demonstrate the capability of the refined PBT model by comparing model predicted pharmacokinetics of ET and EO with published data in mouse, rat and human,
3) demonstrate model capability to predict ET- or EO-induced haemoglobin (Hb) or DNA adduct formation with published data in mice, rats and humans, and
4) exemplify the use of the new PBT model as a basis for species-specific risk-estimation by calculating equivalent ET and EO exposures of mice, rats and human with respect to the species-specific Hb and DNA adduct levels.
GLP compliance:
not specified

Test material

Constituent 1
Chemical structure
Reference substance name:
EC Number:
EC Name:
Cas Number:
Molecular formula:
Details on test material:
Ethylene, purity >99.95% was supplied by Messer Griesheim, Dusseldorf, FRG.


Type of population:
Ethical approval:
not specified
Route of exposure:
Reason of exposure:
Exposure assessment:

Results and discussion

Any other information on results incl. tables

There is a considerable body of data reported describing the parameters obtained by model fitting to experimental data, model parameters obtained for model fitting to humans, model validation and comparisons between the current and previous models. These data are not summarised in this summary. These data are considered important for confirming for the reader the adequacy of the model. The data summarised here will focus on model predicted versus experimentally measured and published results comparing

·      ET exposure levels in rats mice and humans comparing measured and predicted

·      EO exposure levels in rats mice and humans comparing measured and predicted

·      ET exposure in rats, and mice comparing predicted and measured Hb and DNA adducts

·      EO exposure in rats and mice comparing predicted and measured Hb and DNA adducts

·      ET exposure in humans comparing predicted and measured Hb adducts


ET toxicokinetics:Closed chamber inhalation studies with rats, mice and humans where atmosphere concentration of ET were measured over time (mice, 1, 2.8, 10, 32, 80, 310, 1000, 3220 or 11000 ppm for 7h in an exposure chamber with a volume of 2.7 L), rats (36, 108, 310, 11000, 3220, 10600 ppm for 6 h in an exposure chamber of 6.4 L)(Fisler et al., 2015)and humans (5 or 50 ppm for 2 h volume of exposure chamber 12 L)(Fisler et al., 1992, Denk 1990) were compared against the same exposure scenarios using the current model. Measured and predicted ET concentrations in the chambers aligned very well confirming that the model could be used to predict inhalation intake and elimination of ET in rats mice and humans. Neither the measured data nor the model could predict suicide inhibition of ET-metabolising CYP2E1 from concentration curves of atmospheric ET. When the kinetic constant k3 was set to zero (kinetic model above) so that no EO was formed and all of the enzyme substrate complex resulted in suicide inhibition of ET metabolism, the model very nicely agreed with measured data generated from rats and mice pre-exposed the CYP2E1 inhibitor, sodium diethyldithiocarbamate trihydrate, for 30 min prior to ET exposure, demonstrating that if the model was set correctly, atmospheric ET concentrations could be used to predict this scenario.

Similarly, exposing 4 human volunteers to 5, 20 or 50 ppm ethylene and measuring the atmosphere and exhaled air ET concentrations and using these data to calculate the alveolar retention of ET, the model was in good agreement with measured values

EO toxicokinetics.As with ET, the model reliably predicted EO exposure from measurements of atmospheric EO taken at different time points in closed exposure chambers.

EO concentrations in venous blood and exhaled air:The model also successfully predicted the concentration time course of EO in venous blood collected from male B6C3F1 mice during and after a 4 h inhalation experiment. There data are not presented in this summary and the reader should consult the original publication.

The model also accurately predicted the measured concentration time course of EO in exhaled air from four volunteers.

EO adduct formation following exposure to ET:In the same study(Fisler et al., 2013)the EO concentrations in venous blood from the same volunteers exposed to ET as described above was measured. The PBT model accurately predicted venous blood EO concentrations in these individuals.

Using the model to predict Hb and lymphocyte DNA adduct formation by simulating published experiments revealed that following exposure to ET, on average the model predicted HEV and HEG values of 80% and 91% respectively of measured values. Following exposure to EO, on average the model predicted measured HEV levels in rats (107% of measured levels) and mice (98% of measured levels). Similarly, HEG levels were 112% (rats) and 160% (mice) of measured levels.

Predicted and measured HEV and HEG adduct formation at steady state following occupational exposure:Estimating health risk of ET results from exposure to its metabolite EO. Literature reported HEV levels in humans at steady state inhalation exposures (8 h/day, 5 d/w) to EO were compared to predicted levels were compared and were in agreement. Steady state levels predicted by the PBT model following exposure of a reference man to EO (1 ppm) (8 h/d, 5 d/w) are predicted to be 18 weeks for HEV as erythrocyte lifetime is not influenced by EO. HEG was predicted to reach steady state at 3 weeks due to short half-life of HEG depurination (3.75 days). However, data are lacking to validate the model for these parameters in humans.

PBT model predicted ET exposure versus equivalent EO exposures induce the same level of adducts to Hb and lymphocyte DNA.

To estimate health risk of ET, an understanding of the EO exposure levels which are equivalent to ET exposures with respect to Hb and DNA adduct formation is required. Therefore, the PBT model was used to calculate exposure concentrations of EO in mice rats or humans (after one erythrocyte life span (steady state exposure)) that result in a similar level of Hb and DNA adduct formation that would result from an equally long exposures to ET. Modelling exposure of mice and rats (6 h/d, 5 d/w) to ET concentrations of 40, 1,000 and 3,000 ppm, or humans exposed to 200 (8 h/d, 5 d/w) (the time weighted average threshold limit of ET exposure set by the American Conference of Governmental Industrial

Hygienists) were simulated and compared to the literature(Walker et al., 2000).The results are presented in the table below. Additionally, a mouse, rat or human exposure of 10,000 ppm was modelled to predict equivalent EO concentrations that would result similar Hb and DNA adduct levels. Results predict that exposures to 10,000 ppm ET would be required to induce the same adduct levels as EO exposure to mice (3.95 ppm), rats (5.67 ppm) or humans (0.313 ppm).

Applicant's summary and conclusion

The present model for inhaled ET and EO in mouse rat and humans improves upon existing models and quantitively conveys the effects of EO concentrations in blood that result from CYP2E1-catalysed metabolism of ET, ET induced suicide inhibition of CYP2E1, incorporating the rate of CYP2E1 turnover. It further includes parameters that account for EO kinetics with respect to metabolism by EHs and GSTs. The model accurately predicts blood EO concentrations following exposure to ET and estimates Hb and lymphocyte DNA adduct formation in rats mice and humans, thereby confirming model applicability to assess health risks of ET or EO exposure.
Executive summary:

Ethylene (ET) is the largest volume organic chemical. Mammals metabolize the olefin to ethylene oxide (EO), another important industrial chemical. The epoxide alkylates macromolecules and has mutagenic and carcinogenic properties. In order to estimate the EO burden in mice, rats, and humans resulting from inhalation exposure to gaseous ET or EO, a physiological toxicokinetic model was developed. It consists of the compartments lung, richly perfused tissues, kidneys, muscle, fat, arterial blood, venous blood, and liver containing the sub-compartment endoplasmic reticulum. Modeled ET metabolism is mediated by hepatic cytochrome P450 2E1, EO metabolism by hepatic microsomal epoxide hydrolase or cytosolic glutathione S-transferase in various tissues. EO is also spontaneously hydrolyzed or conjugated with glutathione. The model was validated on experimental data collected in mice, rats, and humans. Modeled were uptake by inhalation, wash-in–wash-out effect in the upper respiratory airways, distribution into tissues and organs, elimination via exhalation and metabolism, and formation of 2-hydroxyethyl adducts with hemoglobin and DNA. Simulated concentration-time courses of ET or EO in inhaled (gas uptake studies) or exhaled air, and of EO in blood during exposures to ET or EO agreed excellently with measured data. Predicted levels of adducts with DNA and hemoglobin, induced by ET or EO, agreed with reported levels. Exposures to 10000 ppm ET were predicted to induce the same adduct levels as EO exposures to 3.95 (mice), 5.67 (rats), or 0.313 ppm (humans). The model is concluded to be applicable for assessing health risks from inhalation exposure to ET or EO.