Registration Dossier

Diss Factsheets

Administrative data

Workers - Hazard via inhalation route

Systemic effects

Long term exposure
Hazard assessment conclusion:
DMEL (Derived Minimum Effect Level)
Value:
2.21 mg/m³
Most sensitive endpoint:
carcinogenicity
Route of original study:
By inhalation
DNEL related information
DNEL derivation method:
other: Cox regression (see endpoint summary for details)
Overall assessment factor (AF):
1
Modified dose descriptor starting point:
other: A Cox regression model for leukaemia, reported by Cheng et al (2007).
Explanation for the modification of the dose descriptor starting point:
No route-to-route extrapolation necessary
AF for dose response relationship:
1
Justification:
accounted for in the model
AF for differences in duration of exposure:
1
Justification:
accounted for in the model
AF for interspecies differences (allometric scaling):
1
Justification:
model derived from human epidemiology studies
AF for other interspecies differences:
1
Justification:
model derived from human epidemiology studies
AF for intraspecies differences:
1
Justification:
accounted for in the model
AF for the quality of the whole database:
1
Justification:
database is robust
AF for remaining uncertainties:
1
Justification:
no remaining uncertainties
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

Local effects

Long term exposure
Hazard assessment conclusion:
no hazard identified
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

Workers - Hazard via dermal route

Systemic effects

Long term exposure
Hazard assessment conclusion:
high hazard (no threshold derived)
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

Local effects

Long term exposure
Hazard assessment conclusion:
no hazard identified
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified

Workers - Hazard for the eyes

Local effects

Hazard assessment conclusion:
no hazard identified

Additional information - workers

Compositional information

These hydrocarbon streams meet the regulatory definition of UVCB substances, with inherent variations in composition present due to differences in manufacturing history. This variability is documented in the Category Justification, which lists the chemical marker substances present along with an indicative concentration range for each e. g.

Propene 0 – 100%

1,3-Butadiene 0 - <1%

Benzene 0.1 - <1%

Carbon monoxide 0 - <2%

Ethane 0 - <70%

Ethylene 0 - <100%

Propane 0 - <98%

Methane 0 - <99%

Butenes 0 <31%

Butane 0 - < 100%

Isobutane 0 - <56%

2 -methylpropene 0 - <17%

 

Uses:

These hydrocarbon streams are used as intermediates and monomers, hence risk characterization will focus on workers only. (The presence of >0.1% butadiene and >0.1% benzene precludes their supply to the general population.)

Substance selection for risk characterization:

See "Selection of constituents for HH exposure" in section 13.2.

Benzene

Benzene causes adverse effects on the haematopoietic system of animals and in humans after repeated dose exposure via oral or inhalation routes. Long term experimental carcinogenicity bioassays have shown that it is a carcinogen producing a variety of tumours in animals (including lymphomas and leukaemia). Human epidemiological studiesprovide clear and consistent evidence of a causal association between benzene exposure and acute myelogenous (non-lymphocytic) leukemia (or ANLL). An effect on bone marrow leading to subsequent changes in human blood cell populations is believed underpin this response.

 

The DMEL used in the original versions of the REACH category K dossier was based on the EU BOELV of 1 ppm which was derived from the position on benzene toxicology presented by SCOEL in SUM 140 (SCOEL, 1991). Our analysis of the body of research that has developed since then agrees with the conclusion of DECOS (Netherlands) that the evidence on benzene justifies the setting of a DNEL rather than a DMEL (DECOS, 2014). This position is based on the view that benzene is not a direct-acting mutagen, that clastogenic events will have a threshold and that the key toxicity is haematotoxicity. If haematotoxicity is avoided, then progression to oncological disease would not be expected (LOA 2017).

The use of the EU BOELV as a basis for a DMEL was based on the provision in REACH guidance that allows a DNEL/DMEL to be based on accepted formal workplace limits providing that no data exist that would contradict the basis of the formal workplace limit. (ECHA Guidance R8 Appendix 13). Pending the setting of a new EU BOELV value for benzene, LOA believes that the DECOS document and other recent literature provide enough justification to contradict the 1 ppm 8h TWA EU BOELV. As an interim position LOA previously saw that haematological data reviewed by the DECOS, as well as more recent research provided justification for a DNEL of 0.6 ppm as an 8h TWA.

 

In 2017, ECHA’s Risk Assessment Committee (RAC) was tasked with providing an Opinion on a Benzene OEL. This was provided in March 2018 and proposed an OEL of 0.05 ppm as an 8h TWA. RAC also believed that benzene could be seen as a threshold carcinogen, where avoidance of structural and numerical chromosomal aberrations and micronuclei would protect against cancer risk. (ECHA 2018) During and subsequent to this RAC review of the benzene OEL by the Risk Assessment Committee of ECHA (RAC), LOA have reassessed the data on benzene in greater detail.

LOA believe that the available data show that an OEL of 0.25 ppm/8hTWA is sufficient to protect all aspects of worker health (i.e. cancer, haematological and genotoxic effects). The protection for carcinogenic effects is driven by the evidence for benzene having a thresholded mode of action of cancer, thus the OEL would protect against benzene induced cancer (i.e. Acute Myeloid Leukemia).

Note that Registrants referring to a DNEL of 0.25 ppm (8h TWA) will still be subject to the requirements of the Carcinogens and Mutagens Directive (Council Directive 1999/38/EC as amended) which requires substitution where feasible, exposure minimisation and monitoring of workers. (For references see section 13 "Worker DNEL Explanation").”

Acute toxicity - systemic effects

Benzene is legally classified as cancer Cat 1A in the EU. According to ECHA Guidance Part E: Risk Characterisation, Version 3.0 the substance belongs to the high hazard category, no threshold was derived as the substance is only used as intermediate.

Acute toxicity – local effects

The risk management measures in place for carcinogenicity will provide adequate protection against the occurrence of local effects following acute exposure.

Sensitization

No hazard with respect to skin or respiratory sensitization has been identified in animal studies or in humans and, consequently, no DNEL can or will be proposed.

Oral

The oral route is not relevant to workers and a DN(M)EL will not be proposed.

Dermal

The risk management measures in place for carcinogenicity will provide adequate protection against the occurrence of dermal effects.

Inhalation

The risk management measures in place for carcinogenicity will provide adequate protection against the occurrence of inhalation effects.

Long-term local effects

The (interim) systemic long term threshold and risk management measures in place for carcinogenicity will provide adequate protection against the occurrence of local long-term effects.

  

1,3-Butadiene

A DNEL for acute effects should be derived if an acute hazard leading to acute toxicity (eg. C&L) has been identified and there is a potential for high peak exposures. This is not the case with 1,3-butadiene.

 

The DMEL for long term systemic effects was selected from the most sensitive endpoint (carcinogenicity). DNELs for repeat-dose toxicity (inhalation exposure) and reproductive toxicity (developmental toxicity) were 371 mg/m3and 15 mg/m3respectively. The DMEL of 2.21 mg/m3for carcinogenicity is equivalent to 1 ppm. No long-term local effects have been reported.

 

In experimental animals, there is a marked species difference in carcinogenicity (EU RAR 2002). In the mouse, 1,3-butadiene is a potent multi-organ carcinogen. Tumours develop after short durations of exposure, at low exposure concentrations and the carcinogenic response includes rare types of tumours (NTP 1993). In the rat, fewer tumour types, mostly benign, develop at exposure concentrations of 100 to1000-times higher (Owen et al 1987). In humans a positive association was demonstrated between workplace exposure to butadiene for men employed in the styrene-butadiene rubber industry and lymphohaematopoietic cancer (leukemia) (Sathiakumar et al 2005, Graff et al 2005, Delzell et al 2006, Cheng et al 2007, Sielken et al 2006, 2007 & 2008, Sielken and Valdez-Flores 2013)

 

Worker – long-term systemic inhalation DN(M)EL

The association between 1,3-butadiene exposure and leukemia has been extensively modeled. The excess risk of leukemia as a result of exposure to 1,3-butadiene has then been determined from these models. The details of this approach can be found in the Summary and Discussion of Carcinogenicity Section. The preferred model for workers is the Cox continuous model adjusted only for age (Cheng et al, 2007)based on all leukemias combined, using the exposure metric that excluded exposure that occurred more than 40 yearsago. This model incorporates dose descriptors and assessment factors and therefore further corrected dose descriptors and overall assessment factors are not required. Risk estimates were calculated using mortality rates and lifetables for 2008-10 for the 28 states of the EU. The estimate of the excess risk of death from leukemia as a result of exposure to a DMEL of 2.21 mg/m3(1 ppm) is 0.39 x 10-4.

 

1,3-Butadiene is a gas at room temperature and therefore exposure by the dermal route is not relevant.

 

Carbon monoxide

 

Worker – long-term systemic inhalation DN(M)EL

An IOELV has not been established for carbon monoxide, however, the toxicology and human health hazard has been comprehensively assessed by WHO (1999). The WHO air quality guidance value of 10 mg/m3, intended to ensure that a carboxy-haemoglobin (COHb) level of 2.5% is not exceeded when a normal subject engages in light or moderate exercise, has been considered as equivalent to the long-term systemic inhalation DNEL required by REACH

 

Carbon monoxide is a gas at room temperature and therefore exposure by the dermal route is not relevant.

 

Propene

Worker – long-term systemic inhalation DN(M)EL

As no adverse systemic effects were reported at the highest dose level tested in a chronic toxicity study, no DNELs have been derived for propene.

The weight of evidence indicates that local nasal effects recorded in rodents, which unlike man are obligate nasal breathers, superimposed on a high background of spontaneous nasal pathology and with no obvious dose-response relationships, are likely to be of little relevance in extrapolation of risk to humans and considered an inappropriate finding from which to derive a DNEL. In addition, the concentrations at which these effects were reported in the animal experiments are very high compared to the actual human exposure levels in practice.

Dermal and oral studies with propene are not technically feasible as this substance is a gas at room temperature.

Methane, ethane, butane and propane

No long term systemic DNELs are available for these simple alkanes, while Annex VI of the CLP Regulation records no potential to cause harm following repeated exposure. These component substances are therefore considered to possess low systemic toxicity, and no DNELs are therefore required.

 

Ethylene

Ethylene has been shown to cause subtle nasal effects (rhinitis) in rats in repeated high-exposure inhalation studies. These (and similar) effects have not been reported in humans. The relevance of the rat findings for humans is therefore questionable. Based upon the current animal data, a health-based reference concentration of approximately 50-75 ppm is suggested, i.e., significantly greater than current worker exposure levels. In the circumstances, a DNEL for ethylene is not being proposed. Whilst ethylene has been shown to have anaesthetic properties (at concentrations of 80%, equivalent dose 800,000 ppm or 917,857 mg/m3), the effect is at concentrations which far exceed occupational exposure levels and a clear NOAEC of 10,000ppm has been identified in studies in rats, therefore this should not drive DNEL derivation.

 

Butenes

Repeat dosing studies via inhalation exposure are available for but-1-ene, 2-butene and 2-methylpropene. All studies have shown minimal systemic or target organ toxicity.

 

Exposure of rats to but-1-ene at concentrations of 500, 2000, 8000 ppm (1147, 4589, 18,359 mg/m3) did not induce systemic toxicity in males or females exposed for a minimum of 28 days or in pregnant female rats exposed for 14 days pre-mating, through mating and gestation to day 19. No treatment-related effects on body weight, clinical chemistry, organ weights or histopathology were found. Neurotoxicity screening also showed no effects on motor activity or functional observation battery. A NOAEC of 8000 ppm (18,359 mg/m3) (the highest dose level) was established (Huntingdon, 2003). 

 

Exposure of rats to 2-butene at target concentrations of 2500 or 5000 ppm (5737 or 11,474 mg/m3) did not induce significant systemic toxicity in males and females exposed for 28 days, or in pregnant female rats exposed for 14 days pre-mating, through mating and gestation to day 19 (TNO 1992b). Mean absolute organ weights and relative weights were comparable in all groups. No abnormal, treatment-related macroscopic changes (all groups) or pathological changes (only determined in control and 5000 ppm groups) were observed. The only treatment-related changes were some small decreases in body weights and body weight gains in both sexes at both dose levels and decreased food consumption at 5000 ppm during the first week (premating). Although the authors (TNO 1992b) interpreted the NOAEC as 2500 ppm based on these findings, a reanalysis by RIVM (2007) concluded that as these effects were not dose-related and not consistently present during the study the NOAEC for 2-butene should be 5000 ppm (11,474 mg/m3) (RIVM 2007 and amended SIDS report 2007).

 

2-Methylpropene also caused no toxicologically significant changes when rats were exposed to 250, 1000 or 8000ppm (573, 2294 or 18,359 mg/m3) for 13 weeks. The only clinical change was an elevation in ketone bodies detected in urine at 1000 ppm and 8000 ppm (males), the toxicological significance of this is unknown. The NOAEC was 8000 ppm (18,359 mg/m3) the highest concentration level tested (Hazleton 1982). Similar results were obtained in 14 week inhalation studies conducted by the NTP (NTP, 1998). F344/N rats and B6C3F1 mice were exposed to 2-methylpropene at concentrations of 0, 500, 1,000, 2,000, 4,000, or 8,000 ppm, (1147, 2294, 4589, 9179, 18,359 mg/m3) for 14 weeks. There were no significant exposure-related toxicologic effects in either species at any dose level. Increased kidney weights in mice; and increased liver and kidney weights and minimal hypertrophy of goblet cells lining the nasopharyngeal ducts in rats, were considered to be non-toxic adaptive responses. The NOAEL for both studies was 8000 ppm (18,359 mg/m3) the highest concentration level tested.

 

The absence of significant toxicological findings in these investigations indicates that no repeated dose systemic DNEL is necessary.

Carbon dioxide

No long term systemic DNEL is available for carbon dioxide, while classifications notified to the CLP inventory indicate no potential to cause long-term systemic effects.

 

 

References

 

Blank IH, McAuliffe DJ (1985). Penetration of benzene through human skin. J. Invest. Dermatol. 85, 522–526.

Committee on Hazardous Substances (2008). Guide for the quantification of cancer risk figures after exposure to carcinogenic hazardous substances for establishing limit values at the workplace. 1. Edition.: Bundesanstalt für Arbeitsschutz und Arbeitsmedizin. Availablehttp://www.baua.de/nn_21712/en/Publications/Expert-Papers/Gd34,xv=vt.pdf
Cheng H, Sathiakumar N, Graff J, Matthews R, Delzell E (2007). 1,3-Butadiene and leukemia among synthetic rubber industry workers: exposure-response relationships. Chem Biol Interact, 166,15-24.

Maibach HI, Anjo DM (1981). Percutaneous penetration of benzene and benzene contained in solvents used in the rubber industry. Arch. Environ. Health 36, 256–260.

Sielken RL, Valdez-Flores C, Delzell E (2008). Quantitative Risk Assessment of Exposures to Butadiene in European Union Occupational Settings Based on the University of Alabama at Birmingham Epidemiological Study: All Leukemia, Acute Myelogenous Leukemia, Chronic Lymphocytic Leukemia, and Chronic Myelogenous Leukemia. Unpublished report to Lower Olefins Sector Group,,.
Tsuruta H (1996). Skin absorption of solvent mixtures-effect of vehicle on skin absorption of toluene. Ind. Health 34, 369–378.

World Health Organisation (1999). Environmental Health Criteria 213 (Carbon Monoxide, second edition) 1999, updated 2004

General Population - Hazard via inhalation route

Systemic effects

Long term exposure
Hazard assessment conclusion:
DMEL (Derived Minimum Effect Level)
Value:
0.265 mg/m³
Most sensitive endpoint:
carcinogenicity
Route of original study:
By inhalation
DNEL related information
DNEL derivation method:
other: Cox regression (see endpoint summary for details)
Overall assessment factor (AF):
1
Modified dose descriptor starting point:
other: A Cox regression model for leukaemia, reported by Cheng et al (2007).
Explanation for the modification of the dose descriptor starting point:
No route-to-route extrapolation necessary
AF for dose response relationship:
1
Justification:
accounted for in the model
AF for differences in duration of exposure:
1
Justification:
accounted for in the model
AF for interspecies differences (allometric scaling):
1
Justification:
model derived from human epidemiology studies
AF for other interspecies differences:
1
Justification:
model derived from human epidemiology studies
AF for intraspecies differences:
1
Justification:
accounted for in the model
AF for the quality of the whole database:
1
Justification:
database is robust
AF for remaining uncertainties:
1
Justification:
no remaining uncertainties
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

Local effects

Long term exposure
Hazard assessment conclusion:
no hazard identified
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

General Population - Hazard via dermal route

Systemic effects

Long term exposure
Hazard assessment conclusion:
hazard unknown but no further hazard information necessary as no exposure expected
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

Local effects

Long term exposure
Hazard assessment conclusion:
no hazard identified
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified

General Population - Hazard via oral route

Systemic effects

Long term exposure
Hazard assessment conclusion:
DMEL (Derived Minimum Effect Level)
Value:
0.464 µg/kg bw/day
Most sensitive endpoint:
carcinogenicity
Route of original study:
By inhalation
DNEL related information
DNEL derivation method:
ECHA REACH Guidance
Overall assessment factor (AF):
1
Modified dose descriptor starting point:
other: The inhalatory DMEL (ug/m3) was converted into a human oral DMEL (ug/kg bwt/d) by adjusting for differences in uptake between the two routes of exposure (REACH Guidance, Appendix R.8-2, Example B.4).
Explanation for the modification of the dose descriptor starting point:

The inhalatory DMEL (ug/m3) was converted into a human oral DMEL (ug/kg bwt/d) by adjusting for differences in uptake between the two routes of exposure (REACH Guidance, Appendix R.8-2, Example B.4)

AF for dose response relationship:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for differences in duration of exposure:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for interspecies differences (allometric scaling):
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for other interspecies differences:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for intraspecies differences:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for the quality of the whole database:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
AF for remaining uncertainties:
1
Justification:
Excess risk estimates based on a human multiplicative risk, linear in cumulative exposure model (Crump and Allen, 1984), were used as the starting point. No additional assessment factors have been applied given the conservative nature of model, which is based on human lifetime exposure
Acute/short term exposure
Hazard assessment conclusion:
no hazard identified
DNEL related information

General Population - Hazard for the eyes

Local effects

Hazard assessment conclusion:
no hazard identified

Additional information - General Population

The main use of hydrocarbon streams are as intermediates and monomers, hence no exposure of the general population is likely. The presence of >0.1% butadiene and >0.1% benzene precludes the supply of these streams to the general population, however general population DN(M)ELs for the inhalation (butadiene) and oral (benzene) routes have been developed for the assessment of risks to man exposed via the environment.

1,3-Butadiene: General Population – long-term systemic inhalation DN(M)EL

The association between 1,3-butadiene exposure and leukemia has been extensively modeled. The excess risk of leukemia as a result of exposure to 1,3-butadiene has then been determined from these models. The details of this approach can be found in the Summary and Discussion of Carcinogenicity Section. The preferred model for the general population is one used for occupational exposure calculations i.e. the Cox continuous model adjusted only for age (Cheng et al, 2007)based on all leukemias combined, using the exposure metric that excluded exposure that occurred more than 40 years ago. The model incorporates dose descriptors and assessment factors and therefore further corrected dose descriptors and overall assessment factors are not required. Instead, an estimate of the concentration of 1,3-butadiene giving an excess risk of death from leukemia (all cell types combined) of 1 in 105was determined and while a higher value could have been proposed by including BD HITS, a DMEL of 120 ppb (0.265 mg/m3) is proposed.

A long-term inhalation DN(M)EL for 1,3 butadiene of 0.265 mg/m3will therefore be used for an assessment of risks to man exposed via the environment.

References

Cheng H, Sathiakumar N, Graff J, Matthews R, Delzell E (2007). 1,3-Butadiene and leukemia among synthetic rubber industry workers: exposure-response relationships. Chem Biol Interact, 166,15-24.

Benzene: General Population – long-term systemic oral DN(M)EL

Development of inhalation DN(M)EL

Epidemiology studies provide clear and consistent evidence of a causal association between benzene exposure and acute myelogenous (non-lymphocytic) leukaemia (AML or ANLL). IARC (Baan et al., 2009) has recently concluded that, although there is “sufficient” evidence for an increased risk of AML/ANLL in humans, there is only “limited” or “inadequate” evidence of carcinogenicity in humans for other types of leukaemia. An effect of benzene on bone marrow leading to subsequent changes in human blood cell populations is believed to underpin this response. The long-term systemic DN(M)EL for benzene will therefore be based upon the following information:

Human chronic toxicity (Schnatter et al., 2010): NOAEC = 11.18 mg/m3

Human carcinogenicity (Crump, 1994; WHO, 2000; TCEQ, 2007) = 3.25 µg/m3.

The value that is proposed is based on the approach used by WHO (2000) which combined estimates of excess risk for leukaemia calculated by Crump (1994) for four models into a geometric mean estimate. The same four models were used for the derivation of this DMEL but estimates of excess risk for acute myelogenous or acute monocytic leukaemia (AMML) calculated by Crump (1994) were used instead of those for leukaemia. For three of the four models, excess risk estimates calculated by Crump (1994) were used. A more recent estimate of excess risk was available for one model (TCEQ, 2007) and this was used instead of the estimate calculated by Crump (1994). The value of 3.25 µg/m3 (1 ppb) is protective against haematotoxicity, genotoxicity and carcinogenicity and results in a geometric mean excess lifetime risk of AMLL of 0.9 x 10-5.

While information regarding the NOAEC for effects on human bone marrow post-date WHO (2000), a DNEL based on these bone marrow (threshold) findings would be higher (and hence offer less protection) than one based on AMML. It is also the case that it is not possible to ascribe precise concentrations of benzene to the occurrence of human myelodysplastic syndrome, precluding use of this information for development of a DN(M)EL.

As a consequence, an inhalation DMEL for benzene of 1.0 ppb (3.25 µg/m3) is proposed. This value is lower than the air quality limits of 10 µg/m3 and 5 µg/m3 that were established for benzene in subsequent European Directives 2000/69/EC and 2008/50/EC, respectively.

Extrapolation from inhalation DN(M)EL to oral DN(M)EL

Dose descriptor: The inhalation DMEL of 3.25 µg/m3 will be used.

Modification of dose descriptor: Correct the inhalation DMEL to an oral NOAEL (mg/kg/day) by converting the dose absorbed after inhalation into a systemic dose, assuming 50% uptake by the lung and 100% uptake from the GI tract, a sRV24 -hour of 20 m3 and body weight of 70 kg (REACH TGD, Appendix R.8 -2):

Oral NOAEL = [AQS x sRV24 -hour x [50/100]] / body weight

= 3.25 x 20 x 0.5 / 70 = 0.464 µg/kg bw/d

Assessment factors: As the inhalation DMEL is based on general population life-time exposure no assessment factor is needed.

DN(M)ELl-t oral = 0.464 µg/kg bw/d

References

Baan R et al. (2009). A review of human carcinogens - Part F: Chemical agents and related occupations.The Lancet Oncology, 10(12):1143–1144.

Crump KS (1994). Risk of benzene-induced leukemia: a sensitivity analysis of the Pliofilm cohort with additional follow-up and new exposure estimates. J Toxicol Environ Health 42, 219-242.

TCEQ (2007). Texas Commission on Environmental Quality. Development Support Document. Benzene. Chief Engineer’s Office. Available: http: //tceq. com/assets/public/implementation/tox/dsd/final/benzene_71-43-2_final_10-15-07.pdf

WHO (2000) Air Quality Guidelines for Europe, Second Edition. WHO regional publications, European series; No. 91.