HYDROCARBON SOLVENTS PRODUCERS · PDF fileHSPA recommends an SSV equivalent to the existing TRGS 900 and MAK values. Available data shows no effects following exposures to hexane isomers

Chemistry making a world of difference European Chemical Industry Council Avenue E. van Nieuwenhuyse 4 B - 1160 Brussels Belgium Tel: +32 2 676 72 11 Fax: +32 2 676 73 01 [email protected] www.cefic.org

HYDROCARBON SOLVENTS PRODUCERS ASSOCIATION




Hydrocarbon Solvents Producers Association (HSPA) Background Documentation in

Support of RCP Proposal

Background

The Reciprocal Calculation Procedure (RCP) was developed as a framework to provide

consistent and scientifically sound occupational exposure advice for hydrocarbon solvents.

As a basis for the calculation, similar constituents were grouped, and “guidance values”,

based on available data and existing national regulatory values, were recommended for use in

the calculation. The toxicology of individual hydrocarbon constituents has been studied in

comparison with toxicity studies of complex hydrocarbon solvents, with minimal to no

differences observed (summarized in McKee et al, 2015). In other words, the similarities in

the physico/chemical, toxicokinetic, and metabolic properties of defined groups of

hydrocarbon constituents ensures that the potential for interactive effects having undue

influence on the toxicity of complex solvents is of little to no toxicological relevance. Hence,

it is possible to characterize the toxicity of a complex hydrocarbon solvent either on the basis

of its constituents, or in a more generic way, using data from studies of representative

complex solvents.

Our understanding of the toxicology of hydrocarbon solvents has largely remained unchanged

for decades. This is underscored by the fact that with the exception of a few hydrocarbons,

national regulatory advice on hydrocarbon constituents has remained unchanged over the last

30 years. Although new toxicological information on some substances has recently been

published, they are primarily used to validate the already established safe limits for humans.

As an example, (Juran et al, 2014) assessed whether exposure to regular white spirit1 (RWS)

at the current HSPA OEL of 300 mg/m3 was associated with acute CNS depression. This was

accomplished by exposing human volunteers to 100 or 300 mg/m3 of a dearomatized white

spirit2 (DWS) or RWS for 4 hours

3. Overall, the authors concluded that the “300 mg/m

3 (8-hr

TWA) suggested by SCOEL are adequately protective against acute neurotoxic effects”. This

study, along with other human observational studies previously published by Ernstgard et al

(2009a, b) and (Pedersen & Cohr, 1984), support the RCP-derived OEL of 300 mg/m3 for

RWS and the 600 mg/m3 group guidance value (GGV) adopted by TRGS 900 for aliphatic

hydrocarbons in the C9-C14 range (essentially DWS). Several other human validation studies

have been conducted showing the utility of RCP-derived OELs, and are now summarized in a

comprehensive toxicological review of hydrocarbon solvents (McKee et al, 2015).

In the next few sections, the RCP group guidance values (GGVs) are briefly discussed for

each hydrocarbon solvent category in context of existing national regulatory values and

existing published information on constituents and/or complex solvents spanning each

category. To reiterate, it is important to keep some important points in mind – that the RCP as

a method, is based on the additivity principle (i.e. complex solvent effects are the sum of the

individual constituent effects), hence constituent interactions in complex substances is not

expected to be a factor in the toxic response. With that in mind, it is expected that the toxic

response of individual constituents are not dissimilar from those of the complex solvents in

1 Hydrocarbons C9-C14 aliphatics, 20-25% aromatics

2 Hydrocarbons C9-C14 aliphatics, < 2% aromatics

3 Acute neurotoxicity is considered the most sensitive endpoint in humans for aliphatic/aromatic

hydrocarbons within the C9-C15 range.

2

which they are present and can be used to predict the toxic responses of those complex

solvents. This is robustly supported by the available toxicology data and similarities in

physico/chemical and toxicokinetic properties among individual constituents represented by a

group guidance value. In essence, there is little rationale to justify setting different exposure

limit values for complex solvents when the RCP framework is already designed to ensure that

the regulatory limit values already available for the individual constituents present in the

complex solvents are never exceeded.

C5 Aliphatic Hydrocarbons

Background

Figure 1: Graphical illustration of existing regulatory values and REACH DNELs for

pentanes compared to the with the current HSPA group guidance value for C5-C8 aliphatics.

Note the change in ACGIH TLV for pentanes.

As shown in the figure above, the HSPA group guidance value (GGV) for C5-C8 aliphatics is

conservative with regard to existing regulatory and REACH values for all isomers of

pentanes. This value is also consistent with studies that have shown no systemic and acute

central nervous system (CNS) effects with exposure to n-pentane up to 20,317 mg/m3 and

20,000 mg/m3 respectively (Kim et al, 2012; McKee et al., Lammers et al, 2011). Note that

the highest exposure levels tested were approximately half of the lower explosive limits

(LEL) for these solvents.

General conclusions

While it may be scientifically valid to consider pentanes separately from the C5-C8 GGV, the

HSPA considers it more prudent to maintain pentanes within the GGV to limit the number of

excluded constituents to a necessary minimum. It is recommended that the current HSPA

GGV of 1500 mg/m3 be maintained with regard to pentanes.

3

C6 aliphatic hydrocarbons

Background

Figure 2: Graphical illustration of existing German regulatory values and ACGIH TLV-TWA

for hexane isomers (excluding n-hexane and cyclohexane) compared to the with the current

HSPA group guidance value for C5-C8 aliphatics.

As shown in figure 2 above, the existing ACGIH, TRGS 900 and MAK values for isomers of

hexane (excluding n-hexane and cyclohexane) are higher than the current HSPA GGV for C5-

C8 aliphatics. A search of the GESTIS database of international limit values show that the

HSPA GGV is lower than 13 of 18 8-hour limit values reported (range from 1760 – 1800

mg/m3). Lower values were reported for some countries (Sweden for example) but these were

most likely based on inclusion of cyclohexane.

A search of the REACH database for DNELs of single constituents and complex C6

substances, excluding hexane, revealed values in the range 1474 – 5306 mg/m3. These values

are supported by repeated dose studies such as that published by Daughtrey et al. (1999). In

this study, F-344 rats and B6C3F1 mice were exposed by inhalation to 0, 2700, 9000 or

27000 mg/m3 of a commercial hexane solvent containing 52% n-hexane, 15.6%

methylcyclopentane, 27.6% 2- and 3-methylpentane and 3.2% cyclopentane by weight, 6

hours/day, 5 days/week for 2 years. With the exception of an increase in liver tumors in

female mice and histological evidence of irritation of nasal turbinates in all exposed rats, the

authors reported no other toxicologically relevant finding. The finding of increased liver

tumors with lifetime exposures to hydrocarbons in female mice specifically is of doubtful

human relevance. The US National Toxicology Program (NTP) reported similar finding in a

2-year carcinogenicity study of Stoddard solvent (C9-C14 aliphatics; 2-25% aromatics) and

concluded that this was likely related to increased body weight in exposed females. In other

studies, Yang et al. (2014) exposed rats to 880, 3900 or 18000 mg/m3 2-methylcyclopentane

by inhalation for 90-days. Liver weights of exposed rats were statistically significantly

increased with no concomitant change in serum chemistry or histopathological evidence for

4

overt liver toxicity. There was a mild increase in kidney weights in high dose females

although urinalysis parameters were normal, suggesting that the organ weight differences

were not the result of a pathological. Exposure-related signs of irritation (salivation and

rubbing) were observed in both male and female high exposure groups. The authors

concluded that the NOAEC for the substance was 3900 mg/m3.

n-hexane and cyclohexane

It should be noted that the HSPA GGV for C5-C8 aliphatics explicitly excludes n-hexane as

well as cyclohexane (when the latter is present at concentrations > 20%). N-hexane is

excluded by virtue of its unique toxicity (peripheral axonopathy at high exposures) and the

HSPA supports the adoption of the existing MAK/TRGS 900 values for n-hexane (180

mg/m3) as a substance-specific value (SSV) for use in the TRGS 900 adaptation of the RCP.

Cyclohexane receives special treatment because it has a lower SCOEL value (700 mg/m3)

than other C6 aliphatic constituents. The SCOEL value is based on reports of headaches in

some volunteers exposed to 250 ppm (860 mg/m3) for 4 hours (note that all other

neurobehavioral evaluations were negative, even in subjects that reported headaches; i.e.

NOEC > 250 ppm). Note that SCOEL concludes that “this finding cannot be regarded as a

consistently and significantly adverse effect”. The toxicological data (Christoph et al., 2000;

Kreckman et al., 2000; Malley et al., 2000; Lammers et al., 2009) provide evidence that acute

and/or repeated exposure to cyclohexane produces minimal systemic and/or acute CNS

toxicity. Extending the SCOEL value for cyclohexane to cover all other C5-C8 aliphatic

hydrocarbons is not scientifically justified considering there is sufficient data indicating that

the normal/branched aliphatics within the C5-C8 GGV consistently show NOEC values much

higher than the current 1500 mg/m3 GGV. However, to remain consistent with existing

regulatory values, the HSPA has adopted the SCOEL IOELv (700 mg/m3) as an SSV, which

is consistent with the TRGS 900 and MAK values for this substance. In practice, the SSV for

cyclohexane can be ignored if cyclohexane is present at concentrations <20%, i.e. from a

strictly mass fraction point of view, maximum exposure to cyclohexane present at no more

than 20% in solvent is approximately 300 mg/m3

(less than half the HSPA recommended

SSV).

General conclusions

Excepting n-hexane and cyclohexane, the recommended HSPA GGV for C5-C8 aliphatic

hydrocarbons of 1500 mg/m3 is lower than the current ACGIH, TRGS 900 and MAK values

for hexane isomers. For n-hexane, which is excluded from the C5-C8 GGV on the basis of

unique toxic properties, the HSPA recommends accounting for this using the 180 mg/m3

MAK/TRGS 900 value4. For cyclohexane at concentrations >20% in the complex solvent, the

HSPA recommends an SSV equivalent to the existing TRGS 900 and MAK values. Available

data shows no effects following exposures to hexane isomers at concentrations far higher than

current HSPA C5-C8 GGV and SSVs. Hence, there is no justification to change the existing

C5-C8 HSPA GGV with respect to hexane isomers, including SSVs for n-hexane and

cyclohexane.

4 Note that the 180 mg/m

3 value is recommended for the AGS adaptation of the RCP only. HSPA

preferred value for n-hexane, outside of Germany, is the SCOEL TLV of 72 mg/m3

5


Background

Figure 3: Graphical illustration of existing German regulatory values and ACGIH TLV-TWA

for heptane isomers compared to the with the current HSPA group guidance value for C5-C8

aliphatics. Note that “heptane isomers” may in some cases exclude methylcyclohexane.

As shown in figure 3, there is no basis for proposing a change to the HSPA GGV for C5-C8

aliphatics with respect to C7 aliphatics, considering this value is lower than the existing

German, SCOEL and ACGIH values for n-heptane/heptane isomers. The GGV is also

supported by numerous studies of complex C7 aliphatic hydrocarbon substances showing no

effects at concentrations that are several fold higher. As an example, Carpenter et al, (1975)

exposed rats and dogs to 1900, 3700 or 7900 mg/m3 of a “rubber solvent” composed of C6

and C7 aliphatic hydrocarbon constituents, 6 hours/day, 5 days/week for 13 weeks. Overall,

the authors concluded that there were no toxicologically significant effects in the rats and

dogs and considered the NOAEC to be greater than 7900 mg/m3, a value that is 5.3-fold

greater than the current GGV. Similar studies (showing negligible systemic and/or CNS

effects in rodents) comprising aliphatic hydrocarbons within the C7-C9 range are further

summarized in Mckee et al, (2015).

Methylcyclohexane

Compared to other heptane isomers, the MAK value of 810 mg/m3

for methylcyclohexane

(MCH) is slightly greater than half the existing HSPA C5-C8 GGV. In the Netherlands, this

value is even lower, with an HBROEL of 200 mg/m3. Although documentation for the MAK

value was not available for review, the HBROEL documentation was available. The

6

HBROEL recommendation was based on a repeated inhalation toxicity study of MCH in rats,

mice, hamsters and dogs exposed to 1600 or 8000 mg/m3, 6 hours/day, 5/days/week for 12

months (Kinkead et al, 1985). The rats, mice and hamsters were then held for one year

following exposure while dogs were held for five years to assess recovery (Kinkead et al,

1985). Mean body weight changes were observed in male rats (<10%) and male hamsters

(approximately 20%). The only toxicologically relevant lesions reported were kidney effects

that were solely found in 8000 mg/m3 male rats. The cause of the decreased body weights in

the male hamsters is unknown. However, it should be noted that the exposed hamsters rapidly

gained weight and became equivalent to controls within 2 months post exposure, suggesting

this was most likely a secondary effect considering the general lack of significant body

weight changes in other species tested. Although the authors did not specify a NOAEC, they

concluded that their data was supportive of the then existing ACGIH TLV of (1600 mg/m3).

Discounting the reversible body weight changes in hamsters and the male rat-specific kidney

effects, it can be concluded that 8000 mg/m3 MCH had no toxicologically important effects

on rats, mice, hamsters or dogs following 12 months exposure. The DECOS recommendation

(1600 mg/m3 as the starting point in the derivation of the HBROEL) was based on kidney

effects in male rats, a conclusion that did not take into account the lack of human relevance of

this effect unique to male rats (Swenberg & Lehman-McKeeman, 1999; USEPA, 1991).

In other studies, Treon et al. (1943) exposed rabbits to 1162 ppm (4650 mg/m3) MCH, 6

hours/day, 5 days/week for 10 weeks. The authors concluded that the exposures were

“innocuous” to the rabbits. Repeating the same experiment in 1 monkey exposed to 1500

mg/m3 did not cause any notable effects. Treon et al. (1943) exposed 4 rabbits to 2886 ppm

(11,547 mg/m3) MCH, 6 hours/day, 5 days/week for 3 weeks. No adverse effects on the CNS

or systemic toxicity were noted. Slight lethargy was reported in rabbits exposed to 5567 ppm

(22,274 mg/m3) MCH for 6 hrs/day, 5 days/week for 4 weeks (Treon et al, 1943). Mckee et al.

(2011) reported minor and reversible CNS effects in rats exposed to 14000 mg/m3 of a

complex C6/C7 cycloparaffinic solvent, 8 hours/day for 3 consecutive days. The authors

determined that 4200 mg/m3 was a NOEC for CNS effects. Based on published toxicokinetic

data showing that the human NOAEC for CNS effects with cycloparaffins (between the range

C6-C10) is approximately half that of the rat (Hissink et al, 2009), we can conclude that the

human NOAEC for CNS effects is 2100 mg/m3.

General conclusions

In general, with the exception of MCH, the recommended HSPA GGV for C5-C8 aliphatics,

with respect to heptanes and its isomers, is consistent (even lower than) with the existing data

and current German regulatory values for heptanes. The MCH MAK/SCOEL value is

proposed as an SSV in order to remain consistent with the German regulatory value.

However, at concentrations below 40% in complex solvents, it is not necessary to use the

SSV in the calculations as the MAK value would not be exceeded if the C5-C8 GGV is

adopted as the OEL for the complex solvent. For example, strictly applying mass fraction,

the concentration of MCH in ambient air (if the C5-C8 GGV of 1500 mg/m3 is applied as an

OEL) is ≤ 600 mg/m3 when the complex solvent contains ≤ 40% MCH, a value that is less

than the 810 mg/m3 MAK value. As another example, consider a hypothetical hydrocarbon

solvent containing 35% MCH and 65% n-heptane; based on the C5-C8 GGV, the calculated

OEL for this solvent is 1500 mg/m3 (SSV for MCH is ignored since MCH content is < 40% in

solvent mixture). Using Raoult’s law, the maximum ambient air concentration of MCH at

25°C is 538 mg/m3. If it is assumed that MCH levels are approximately 40% in solvent

7

mixture, maximum ambient air concentration (at 1500 mg/m3 OEL) is determined to be 613

mg/m3. In essence, at levels at or below 40% in a complex solvent, MCH vapor concentration

in ambient air is less than 80% of the current MAK value. Hence a need to account for MCH

separately, using an SSV in the RCP process, is not required at MCH levels below 40%.


Background

Figure 4: Graphical comparisons of existing German and US OELs for octane and its isomers

with the current HSPA group guidance value for C5-C8 aliphatics. Note that AGS and MAK

values explicitly exclude all isomers of trimethylpentane. Trimethylpentanes are classified as

3A for carcinogenicity by the DFG.

As shown in the figure above, excluding trimethylpentanes, the HSPA GGV for C5-C8

aliphatics is 1.6-fold more conservative than the AGS and DFG values for n-octane/octane

isomers.

As is documented in published literature, several studies are available which are supportive of

the existing HSPA GGV particularly for octane and its isomers. Sung et al. (2010) conducted

a 13-week inhalation subchronic toxicity study in rats exposed to 0, 930, 2620 or 7480 mg/m3

n-octane, 6 hrs/day, 5 days/week. The authors reported no clinical and histopathological

differences between all groups of rats and determined the NOEC to be greater than 7480

mg/m3. Rats were exposed to 1400, 4200 or 14000 mg/m

3 n-octane via inhalation, 8 hrs/day

for 3 consecutive days and then evaluated for possible CNS effects in a standardized

functional observational battery (FOB). No CNS effects were observed upon exposure and the

NOEC for acute neurotoxicity was determined to be greater than 14000 mg/m3 (Lammers et

al, 2011). In a similar study of a C8 isoparaffin, the NOEC for acute neurotoxicity was

determined to be greater than 14000 mg/m3 (McKee et al, 2011). Schreiner et al. (1998)

conducted a subchronic toxicity study on Sprague-Dawley rats exposed to a light alkylate

naphtha distillate (LAND-2) containing a complex substance of C4-C10 isoparaffins (31% C8

8

fraction in the vapor phase). Rats were exposed to 0, 2400, 8100 or 24300 mg/m3 LAND-2, 6

hours/day, 5 days/week for 13 weeks. Aside from the standard subchronic toxicity parameters

measured, neurotoxicity evaluations were also conducted using a standard FOB during and at

study termination. Aside from the male rat-specific kidney effects, no other treatment-related

effects were observed and the toxicologically relevant NOEC was greater than 24300 mg/m3

(Schreiner et al, 1998).

General conclusions

Based on the available data, there is no subchronic/neurotoxicity concern for n-octane and

octane isomers at the existing HSPA GGV for C5-C8 aliphatics. Trimethylpentanes are

excluded from the AGS and MAK values for octane isomers on the basis of the 3A cancer

classification by the DFG. However, as has been outlined in detail (see Appendix I), the

literature does not provide compelling evidence that the toxicological properties of

trimethylpentanes differ from those of other aliphatic hydrocarbons. In other words, we

propose that the existing HSPA GGV for C5-C8 aliphatic hydrocarbons is protective of

worker exposure and there is no rationale to exclude trimethylpentanes from this value.


Background

No MAK, TRGS 900 or SCOEL values are available for n-nonane and nonane isomers.

According to the TRGS 900 RCP method, a 600 mg/m3 GGV for C9-C15 aliphatics is

provided to cover the C9 aliphatics. It must be noted that this GGV is different to the HSPA

recommended GGV for the C9-C15 aliphatics which is 1200 mg/m3. A search through the

GESTIS international limit values for chemical agents’ database shows a range of OELs

between 1050 and 1200 mg/m3 in 13 of 15 countries. It should be noted that the UK HSE

(which implements a version of the RCP) recommends an OEL of 1200 mg/m3 for all normal

and branched chain alkanes greater than C7 (which includes nonanes). The ACGIH 8-hour

TLV-TWA for n-nonane (1048 mg/m3) was based on a subchronic toxicity study of rats

exposed to 0, 1900, 3100 or 8400 mg/m3 n-nonane, 6 hrs/day, 5 days/week for 13 weeks

(Carpenter et al, 1978). Mean body weights and body weight gain for high dose rats were

statistically significantly lower than control rats at 3, 17, 32, 46 and 61 days after exposure

but the differences were not maintained to the end of the study. Mild transient acute CNS

effects were reported at 8400 mg/m3 and the NOEC was 3100 mg/m

3. The ACGIH concluded

that a TLV-TWA of 1048 mg/m3 would be sufficient to protect against potential CNS

impairment. It should be noted that the ACGIH has proposed this value exclusively for n-

nonane. For other isomers of nonane, the ACGIH recommends a GGV value of 1200 mg/m3

for C9-C15 alkanes under the ACGIH adaptation of the RCP. This value is consistent with the

available data showing acute CNS effects (the critical adverse effect associated with nonane

exposure) occur at considerably higher concentrations. McKee et al, (2011) reported only

minor and transient CNS effects in rats exposed to 5000 mg/m3 of a C9-C11 isoparaffinic

solvent 8 hrs/day for 3 consecutive days, with 1500 mg/m3 as the NOEC. Phillips and Egan

(1984) exposed different rat strains to 1800 or 5300 mg/m3 of a complex C9-C11 isoparaffinic

solvent in two separate 90-day inhalation toxicity studies. Aside from male-specific kidney

effects, the authors did not report any other adverse effects, with the NOAEC determined to

9

be greater than 5400 mg/m3 (highest concentration tested) (Phillips & Egan, 1984a; Phillips &

Egan, 1984b).

A low HBROEL value for n-nonane (500 mg/m3) was found in a search of available

databases. However, it appears that the HBROEL was developed from the same study by

Carpenter et al. (1978) which formed the basis of the ACGIH TLV-TWA for n-nonane.

General conclusions

The current HSPA GGV recommendation of 1200 mg/m3 for C9-C20 aliphatics is consistent

with the UK HSE GGV for aliphatics >C7, the ACGIH GGV for C9-C15 alkanes (excluding

n-nonane which has a TLV-TWA value of 1048 mg/m3) and the OEL values of 1050 – 1200

mg/m3 for 13 other countries on the GESTIS database including Australia, Belgium, Canada,

France, Switzerland and Denmark. This value is also consistent with available data showing

that CNS effects and potential sensory irritation are not likely at lower concentrations. On this

basis, the HSPA believes that the C9-C14 aliphatics GGV of 600 mg/m3, as proposed in

TRGS 900, is sufficiently protective of possible adverse effects.


C10 isomers excluding decalin

No occupational exposure values for n-decane and other C10 isomers could be located in the

GESTIS database except for a Danish 250 mg/m3 OEL for n-decane and 350 mg/m

3 for other

C10 isomers. Documentation justifying the derivation of these values was however not

available for review. On the other hand, the RCP adaptations by the UK HSE and ACGIH

provide a GGV of 1200 mg/m3 for >C7 aliphatics and C9-C15 alkanes respectively. Under

both recommendations, this GGV is expected to include n-decane and other C10 isomers.

This GGV is supported by data from Lammers et al. (2011) showing minimal CNS effects

(but no other adverse effects) in rats exposed to 5000 mg/m3 n-decane, 8hrs/day for 3

consecutive days. The authors concluded that 1500 mg/m3 was the NOEC, the same value

identified as the NOEC in a similar study by Mckee et al. (2011) for a complex C9-C11

isoparaffinic solvent. Carrillo et al, (2013) exposed rats to 2600, 5200 or 10400 mg/m3 of a

C10-C12 isoparaffinic solvent 6 hrs/day, 5 days/week for 13-weeks. The authors reported

statistically significant liver enlargement (without any changes in liver enzyme levels in

serum and histopathological changes indicative of overt liver toxicity) which has been

described earlier as a functional adaptation to increased metabolic load (Ennulat et al, 2010;

Maronpot et al, 2010; Schulte-Hermann, 1974; Schulte-Hermann, 1979). Aside from the

increased liver weights, the only other health effect noted by the authors was male rat-specific

kidney changes which are not relevant for human risk. The NOAEC for this study was

determined to be >10,400 mg/m3. Phillips and Egan (1984a and 1984) concluded that the

NOAEC for rats exposed to a C9-C11 isoparaffinic solvent in a 90-day inhalation toxicity

study was >5300 mg/m3. Amoruso et al. (2008) summarized a series of 90-day inhalation

toxicity studies on C9-C14 aliphatic hydrocarbons published between the 1960s-70s and

found no consistent treatment-related changes other than the male rat-specific kidney effects

10

(Amoruso et al, 2008). These studies are consistent with others published on complex

solvents containing C10 aliphatics that have been reviewed in recent publications (Johnson et

al, 2012; Mullin et al, 1990). Johannsen & Levinskas, (1987) administered 0, 100, 300 or

1000 and 0, 3000, 10000 or 30000 ppm (dietary levels of 300, 1000 or 3000 mg/kg/day) of a

mixture of several isomeric ratios of tetramethylcyclohexane (a C10 cycloparaffin) orally for

90-days to dogs and rats respectively. The authors reported no observable effects on

mortality, behavior, growth, clinical or hematology at all test levels in both species. It was

concluded that the NOAEL was greater than the highest dose tested in both species. Nilsen et

al, (1988) exposed rats to saturated vapor concentrations of a range of normal alkanes (nC9 –

cC13) including n-decane for 8hours, followed by a 14-day observational period. No

behavioral effects were observed with exposure to n-decane, including an evaluation of the

brain (large brain, cerebellar cortex, purkinje cells) at autopsy (NOEC > 7900 mg/m3).

Overall, there is an abundance of data, from single constituents and complex substances,

justifying the conservative nature of the HSPA GGV for C9-C20 aliphatics (1200 mg/m3), a

value that is also recommended by the ACGIH and UK HSE, and the TRGS 900 GGV of 600

mg/m3 for C9-C15 aliphatics.

Decalin

The DFG currently has a MAK value of 29 mg/m3 and a REACH DNEL of 24 mg/m

3 was

also identified for decalin. No document justifying the DFG value could be obtained for

review. In hydrocarbon solvents, decalin is formed via hydrogenation of naphthalene present

in hydrocarbon solvent feedstocks for C9-C15 aliphatics (normal straight-run kerosene).

Since the level of naphthalene in the feedstocks is approximately in the range of 1%, decalin

levels in the finished solvents typically do not exceed 2%. Using Raoult’s law, the maximum

level of decalin in ambient air, assuming a worst-case 2% content level and C9-C15 aliphatics

GGV of 600 mg/m3, is approximately <16 mg/m

3. Since this is approximately half the MAK

value, it is concluded that the MAK value of decalin is never exceeded as long as the 600

mg/m3 GGV is not exceeded.

General conclusions

In general, both the UK HSE and ACGIH adaptations and available data support maintaining

the HSPA GGV of 1200 mg/m3 for C9-C20 aliphatics and the 600 mg/m

3 value recommended

by TRGS 900 for C9-C14 aliphatics in Germany. Due to the low level of decalin in these

solvents, HSPA concludes that a separate accounting for decalin in the RCP method is not

warranted. Furthermore, detailed decalin studies by the NTP (including a 2-year

carcinogenicity study) do not provide evidence for a unique toxicity such as to warrant

exclusion from the broader C9-C20 aliphatics RCP group. With regard to acute toxicity,

Nilsen et al. (1988) provided evidence that the ability to generate significant vapor

concentrations of n-alkanes decreases with increasing molecular weight/carbon number, such

that aliphatic hydrocarbons from >C10 do not generate vapor concentrations sufficient

enough to cause acute CNS effects (see also Table 1 below). Mckee et al. (2010) exposed rats

to a C10 cycloparaffinic solvent 8hrs/day for three consecutive days. No CNS effects were

reported at 5000 mg/m3 (the highest concentration tested). In conclusion, both from a

standpoint of systemic toxicity and acute CNS effects, there does not appear to be any

justification to reduce the GGV for n-decane and decane isomers beyond values already

considered protective in TRGS 900.

11

>C10 Aliphatic Hydrocarbons

Similar to the C10 aliphatic hydrocarbons, regulatory values for aliphatic hydrocarbons

greater than C10 are sparse to non-existent. A search through the GESTIS database revealed

no national values for these substances. However, RCP adaptations by the UK HSE and

ACGIH provide a GGV of 1200 mg/m3 for >C7 aliphatics and C9-C15 alkanes respectively.

These recommendations are consistent with the HSPA GGV for C9-C20 aliphatic substances

of 1200 mg/m3.

Many studies are available to provide support for the current HSPA GGV. However, a

common problem with extrapolating or deriving DNELs/OELs from many of these studies is

that the NOAECs are essentially conservative estimates that are largely dependent on

experimental constraints. In the example shown in the table below, with the exception of

undecane (C11) and dodecane (C12), maximum vapor concentrations of all alkanes ≥ C13 are

less than half the HSPA GGV at 25 °C.

Table 1: Vapor concentrations of n-alkanes

Substance

Initial

boiling

point

(oC)

vapor pressure @

25oC (mm Hg)

vapor

concentration.

@ 25oC (mg/m

3)

C10-C20

aliphatics

GGV

(mg/m3)

n-decane 174 1.3 10000 1200

n-undecane 196 0.42 3500 1200

n-dodecane 216 0.16 1500 1200

n-tridecane 235 0.055 545 1200

n-tetradecane 253 0.019 202 1200

n-pentadecane 271 0.006 68 1200

n-hexadecane 286 0.005 61 1200

n-heptadecane 294 0.0015 20 1200

n-octadecane 309 0.0008 11 1200

n-eicosane 342 0.00014 2 1200

The toxicity of n-C9 to n-C13 alkanes following short term acute exposures were evaluated in

the rat (Nilsen et al. 1988). Saturated vapor concentrations decreased with increasing carbon

number, with n-C9 highest at 5280 ppm (27583 mg/m3), 442 ppm (2800 mg/m

3) for n-C11,

142 ppm (985 mg/m3) for n-C12 and 41 ppm (308 mg/m

3) for n-C13. Since it is impossible to

generate sufficiently high test vapor concentrations for studies, a larger number of studies in

this range are conducted by the oral route. Although the vast majority of the inhalation studies

show relatively little to no adverse effects outside the male rat kidney effects, the use of the

“highest dose tested” (more often than not the highest concentration experimentally

12

achievable) as the point of departure for the derivation of DNELs/OELs tends to yield much

lower values than for more volatile aliphatic hydrocarbons where data on much higher test

concentrations are possible. With the oral studies, the ability to test sufficiently high enough

doses is also limited by experimental constraints, such as the high propensity for aspiration-

related deaths with oral exposures of hydrocarbons and dose limits imposed by standard

regulatory testing. In essence, using these limits (where true NOAELs are greater than highest

dose tested) is not a true reflection of the safe limits for these set of hydrocarbons.

In the 90-day inhalation study by Carrillo et al. (2013) of a C10-C12 isoparaffinic solvent,

male rat-specific kidney effects were the only toxicologically relevant changes noted. The

NOAEC for the study was determined to be 10,400 mg/m3 (the highest concentration tested).

Carrillo et al. (2013) also reported an unpublished 90-day oral toxicity study of a C11-C15

isoparaffinic solvent in rats administered up to 1000 mg/kg bw/day. Reported effects were

similar to the commonly reported effects in other inhalation studies: increased liver weight

with no histopathological or serum chemistry indication of overt liver toxicity, male-rat

specific kidney effects and small statistically significant changes in hematological parameters

that have been shown to be within reference range for these rat strain and age. NOAEL was

determined to be greater than highest dose tested. In an unpublished 90-day study, rats were

exposed to 0, 1500, 3000 or 6000 mg/m3 of a C9-C11 naphthenic solvent (approximately 70%

naphthenes and <2% aromatics), 6 hours/day, 5 days/week. Male-rat kidney effects were the

only adverse effects reported and NOAEC was determined to be greater than highest exposure

concentration. Juran et al (2014) exposed volunteers to 100 or 300 mg/m3 of a dearomatized

white spirit (C9-C12 aliphatics; <2% aromatics) and a regular white spirit (C9-C12 aliphatics;

2-25% aromatics) for 4 hours. The authors reported no neurobehavioral effects in the humans

and concluded that the existing 300 mg/m3 OEL for regular white spirit was adequately

protective of potential CNS effects. Prior studies by Ernstgard et al (2009a, 2009b) showed no

irritative and/or neurobehavioral effects in human volunteers exposed to 600 mg/m3 of the

same dearomatized white spirit. (Pedersen & Cohr, 1984) reported no change in subjective

symptoms when human volunteers were exposed to 1228 mg/m3 of the same dearomatized

white spirit and 610 mg/m3 of a C9-C11 regular white spirit (18% aromatics) for 6 hours.

Aliphatic substances >C15 are generally high boiling, low vapor pressure constituents that are

practically incapable of sufficiently high vapor concentrations as to cause any adverse effects.

In this case, aerosolization appears to be a more critical concern than vapor concentrations.

The HSPA RCP method does not cover aerosol exposures. Rather HSPA recommends that

exposures be kept below maximally attainable vapor concentrations to avoid aerosol

formation. If aerosols are formed, HSPA recommends that the ACGIH and/or SCOEL TLVs

for fine mineral oil mists should be observed.

General conclusions

Based on the available data and practical constraints of aliphatics in this group, it is

recommended that the HSPA GGV of 1200 mg/m3, consistent with the ACGIH and UK HSE

RCP adaptations, or the 600 mg/m3 value under the German AGS adaptation, be maintained

in the absence of any other data to suggest otherwise. This value is mainly protective of the

most volatile ends of the range (C9-C10) and is supported by validation studies in humans

showing no evidence for CNS effects at 1200 mg/m3 for the dearomatized white spirits and

600 mg/m3 for the regular white spirits. Beyond C11, a sufficient vapor concentration to cause

acute CNS effects would not be expected.

13

C7-C8 Aromatic Hydrocarbons

Based on the inconsistencies in regulatory values for individual constituents in the C7-C8

aromatics category, the HSPA recommends the withdrawal of the former GGV of 200 mg/m3

for C7-C8 aromatics. The HSPA is recommending that for substances containing individual

C7-C8 aromatics at levels > 1%, the current occupational exposure limits for the individual

constituent should be used as a specific substance value (SSV).

C9-C15 Aromatic Hydrocarbons

The HSPA recommended GGV for C9-C15 aromatics of 100 mg/m3 is based on existing TLV

and IOELVs for trimethylbenzene isomers and cumene (isopropylbenzene). This

recommendation is consistent with the TRGS 900 recommendation for the same GGV and is

5-fold lower than the UK HSE recommendation. With the exception of the MAK value for

cumene of 50 mg/m3, all other regulatory values available through the GESTIS database

(including ACGIH 2014 and SCOEL IOELV values) for trimethylbenzene and cumene range

from 100-125 mg/m3 and 100 – 246 mg/m

3 respectively. The HSPA recommendation is also

consistent with the ACGIH RCP GGV for C9-C15 aromatics (100 mg/m3). These values are

consistent with the available data such as that published by Clark et al. (1989). In this study,

rats exposed to a blend of C9 aromatic hydrocarbons (predominantly made of

trimethylbenzenes and ethyltoluenes) for 12 months by inhalation, did not show any evidence

to toxicologically relevant effects up to 1800 mg/m3 (highest concentration tested). Clark et

al. (1989) also reported on a 13-week inhalation toxicity study in rats exposed to the same

substance up to 7400 mg/m3. The authors reported an increase in liver and kidney weights in

female rats but were not considered toxicologically relevant due to the lack of

histopathological correlates indicating overt toxicity. A NOAEC for the 90-day toxicity study

was determined to be the highest concentration tested (7400 mg/m3) (Clark et al, 1989).

Acute CNS effects

The human evidence indicates that there are no CNS effects in humans exposed to

concentrations approximating the HSPA recommended GGV. Early human exposure studies

on complex C9 aromatic solvents revealed evidence for acute CNS effects and respiratory

irritation in humans exposed to concentrations between 10 and 60 ppm (approximately 50 and

300 mg/m3) (Battig et al., 1956). Based on these human observational data, early

recommendations for occupational exposure limits for C9 aromatic solvents were in the range

of approximately 35-50 ppm (approximately 175-250 mg/m3) (Carpenter et al., 1975;

Gerarde, 1960; Nau et al., 1966). Volunteer studies in humans showed no evidence for

transient CNS effects or respiratory irritation with exposure to 25 ppm (123 mg/m3) 135- or

124-trimethylbenzene (Jarnberg et al., 1996; Jarnberg et al., 1998; Jarnberg et al., 1997; Jones

et al., 2006) or 150 mg/m3 (~ 30 ppm) 135-trimethylbenzene for 8-hours (Kostrewski and

Wiaderna-Brycht, 1995; Kostrzewski et al., 1997). It should be noted that these human data

form the basis for the current ACGIH TLV for individual trimethylbenzenes. McKee et al,

(2010) exposed rats to a 200, 1000 or 5000 mg/m3 of a complex C9 aromatic solvent or 125,

1250 or 5000 mg/m3 of 1,2,4-trimethylbenzene, 8 hours/day for 3 consecutive days. Mild but

statistically significant effects were noted at 5000 mg/m3 (124-trimethylbenzene) and 1000

14

and 5000 mg/m3 in the complex C9 aromatic solvent. The NOEC for the study was

determined to be 200 mg/m3. Acute exposures of rats to 2500 or 6000 mg/m3 cumene resulted

in an increase in motor activity (Cushman et al, 1995).

CNS effects with Repeated-exposures

In a repeated-exposure study of a complex C9 aromatic solvent to determine the potential for

long term neurological damage, rats were exposed to 500, 2500 or 7500 mg/m3, 6 hours/day,

5 days/week for 13 weeks (Douglas et al, 1993). Neurobehavioral assessments such as a

functional observation battery and assessment of motor activity were conducted at least 24

hours after the last exposure to avoid the confounding effects of transient acute CNS effects.

Neurological tissue was also assessed for the presence of histopathological lesions indicative

of permanent CNS damage. No neurobehavioral and/or neuropathological effects were

observed in these rats. The authors concluded that the NOEC for neurological damage was

greater than the highest concentration tested. In another study, rats were repeated exposed to

cumene vapors up to 6000 mg/m3 (Cushman et al, 1995). The authors reported no effects on

functional observations and auditory brain stem measurements. Similar to the Douglas study,

no histopathological lesions were found in central or peripheral nervous tissue assessed.

Special considerations

Special consideration should be given to substances with unique toxicities such as

diethylbenzene and triethylbenzene which are known to generate chromogenic γ-diketone

metabolites that cause similar peripheral nervous system effects as observed with n-hexane.

HSPA recommends that these substances be accounted for using an SSV. In the absence of

existing European regulatory values, HSPA proposes the use of American Industrial Hygiene

Association (AIHA) 8-hour TWA of 28 mg/m3. For biphenyl which has a low ACGIH TLV

of 1.5 mg/m3, this value can be used as a SSV. However, due to its low vapor pressure and the

fact that HSPA has adopted a 1.5% content limit on biphenyl in complex solvents, ambient air

vapor concentrations are not expected to exceed its TLV even if the 100 mg/m3 GGV was

applied. In that case, for biphenyl levels <1.5%, no SSV is required.

Unlike other alkylated benzenes and alkylated naphthalenes, naphthalene is metabolized

primarily through ring oxidation, which may introduce metabolites with unique toxicological

properties. In the absence of a definitive regulatory value for naphthalene, HSPA proposes to

continue using the 50 mg/m3 historical OEL (based on human observations) as an SSV

pending the completion of ongoing human observational studies in Germany. HSPA supports

the replacement of this value with the final regulatory value as determined by the AGS.

Methylnaphthalene is metabolized through side chain oxidation (80%) and ring oxidation

(20%). In light of the small metabolic difference (compared to alkylated benzenes), it is

proposed that a 50 mg/m3 SSV be considered for this substance in the absence of SCOEL,

TRGS 900 or MAK values. In the alternative, an exposure validation program should be

considered to ensure validity of existing 100 mg/m3 GGV.

General conclusions

There is no new data supporting a need to change the 100 mg/m3 GGV for C9-C15 aromatic

hydrocarbons. HSPA recommends that other aromatic substances with unique toxicology

(diethylbenzene as an example) and metabolic differences that may influence toxicity

(naphthalene) should be accounted for separately using SSVs. In the case of biphenyl the low

15

vapor pressure and low level in complex solvents suggests that its SSV can be ignored in the

context of the C9-C15 aromatics GGV. With regard to cumene, there is a need to

accommodate the lower MAK value of 50 mg/m3 within the GGV of 100 mg/m

3. However, it

should be noted that conservative worst-case levels of cumene in C9 aromatic solvents are

below 10% (likely containing worst-case levels of cumene). As an example, the complex C9

aromatic substance tested by Clark et al, (1989) contained between 0.63 – 2.8% cumene.

Assuming a worst-case 10% level of cumene in a hypothetical C9 aromatic solvent,

maximum levels of cumene in ambient air (at the GGV of 100 mg/m3) is calculated to be 22

mg/m3 (using Raoult’s law). As this value is below half the MAK value, we can conclude that

at current low levels of cumene in hydrocarbon fluids, a need to account for cumene

separately in the RCP is not warranted and that cumene can be accomodated within the

existing C9-C15 aromatics GGV.

Brussels July 22, 2015

Jose A. Ruiz

Chairman, Hydrocarbon Solvents Producers Association (HSPA)

Antoine Brossier

Director General European Solvents Industry Group (ESIG)

References

Amoruso MA, Gamble JF, McKee RH, Rohde AM, Jaques A (2008) Review of the

toxicology of mineral spirits. International journal of toxicology 27: 97-165

API. (2008) Gasoline Blending Streams Analytical Data.

http://www.petroleumhpv.org/DE3BA876-B71D-4BBF-BBB4-

233492F68365/FinalDownload/DownloadId-

1AD76093DC6771F4477CDAEE17E805DF/DE3BA876-B71D-4BBF-BBB4-

233492F68365/petroleum-substances-and-

categories/~/media/PetroleumHPV/Documents/2008_aug21_gasoline_catanalysis_final_analy

tical_data_summary.ashx

Argyris TS (1971) Additive effects of phenobarbital and high protein diet on liver growth in

immature male rats. Dev Biol 25: 293-309

Battig, K., Grandjean, E., Turrian, V., 1956. Damage to health after long-term exposure to

trimethylbenzene in a paint shop. Zeitschrift für Präventivmedizin 1: 389-406.

16

Boone L, Meyer D, Cusick P, Ennulat D, Bolliger AP, Everds N, Meador V, Elliott G, Honor

D, Bounous D, Jordan H (2005) Selection and interpretation of clinical pathology indicators

of hepatic injury in preclinical studies. Vet Clin Pathol 34: 182-188

Carpenter CP, Geary DL, Jr., Myers RC, Nachreiner DJ, Sullivan LJ, King JM (1978)

Petroleum hydrocarbon toxicity studies XVII. Animal response to n-nonane vapor. Toxicol

Appl Pharmacol 44: 53-61

Carpenter, C., Kinkead, E., Geary, D., Sullivan, L., King, J., 1975. Petroleum hydrocarbon

toxicity studies VII. Animal and human response to vapors of “70 solvent”. Toxicology and

Applied Pharmacology 34:395-412.

Carpenter CP, Kinkead ER, Geary DL, Jr., Sullivan LJ, King JM (1975) Petroleum

hydrocarbon toxicity studies. IV. Animal and human response to vapors of rubber solvent.

Toxicol Appl Pharmacol 33: 526-542

Carrillo JC, David Adenuga M, McKee RH, Roth RN, Steup D, Simpson BJ (2013) The sub-

chronic toxicity in rats of isoparaffinic solvents. Regulatory toxicology and pharmacology :

RTP 67: 446-455

Christoph G, Kelly D, Krivanek N (2000). Acute inhalation exposure to cyclohexane and

schedule-controlled operant performance in rats: Comparison to D-amphetamine and

chlorpromazine Drug and Chemical Toxicology 23:539-553.

Clark, D. G., Butterworth, S. T., Martin, J. G., Roderick, H. R., Bird, M. G., 1989. Inhalation

toxicity of high flash aromatic naphtha. Toxicology and Industrial Health 5: 415-428.

Cushman JR, Norris JC, Dodd DE, Darmer KI, Morris CR (1995) Subchronic Inhalation

Toxicity and Neurotoxicity Assessment of Cumene in Fischer 344 Rats. International journal

of toxicology 14: 129-147

Daughtrey W, Newton P, Rhoden R, Kirwin C, Haddock L, Duffy J, Keenan T, Richter W,

Nicolich M (1999) Chronic inhalation carcinogenicity study of commercial hexane solvent in

F-344 rats and B6C3F1 mice. Toxicological sciences : an official journal of the Society of

Toxicology 48: 21-29

DFG (2002) Trimethylpentane (all isomers) [MAK Value Documentation, 2006]. In The

MAK-Collection for Occupational Health and Safety. Wiley-VCH Verlag GmbH & Co.

KGaA

Douglas JF, McKee RH, Cagen SZ, Schmitt SL, Beatty PW, Swanson MS, Schreiner CA,

Ulrich CE, Cockrell BY (1993) A neurotoxicity assessment of high flash aromatic naphtha.

Toxicology and industrial health 9: 1047-1058

Ernstgard L, Iregren A, Juran S, Sjogren B, van Thriel C, Johanson G (2009a) Acute effects

of exposure to vapours of standard and dearomatized white spirits in humans. 2. Irritation and

inflammation. Journal of applied toxicology : JAT 29: 263-274

17

Ernstgard L, Lind B, Johanson G (2009b) Acute effects of exposure to vapours of standard

and dearomatized white spirits in humans. 1. Dose-finding study. Journal of applied

toxicology : JAT 29: 255-262

Ennulat D, Walker D, Clemo F, Magid-Slav M, Ledieu D, Graham M, Botts S, Boone L

(2010) Effects of hepatic drug-metabolizing enzyme induction on clinical pathology

parameters in animals and man. Toxicologic pathology 38: 810-828

Gaworski CL, Haun CC, MacEwen JD, Vernot EH, Bruner RH, Amster RL, Cowan MJ, Jr.

(1985) A 90-day vapor inhalation toxicity study of decalin. Fundamental and applied

toxicology : official journal of the Society of Toxicology 5: 785-793

Gerarde. H., 1960. Toxicology and Biochemistry of Aromatic Hydrocarbons. In: Elsevier

Monographs on Toxic Agents. Browning, E. ed. Elsevier Publishing Company, New York.

Gershbein LL (1958) Pregnancy and liver regeneration in partially hepatectomized rats. Proc

Soc Exp Biol Med 99: 716-717

Hall AP, Elcombe CR, Foster JR, Harada T, Kaufmann W, Knippel A, Kuttler K, Malarkey

DE, Maronpot RR, Nishikawa A, Nolte T, Schulte A, Strauss V, York MJ (2012) Liver

hypertrophy: a review of adaptive (adverse and non-adverse) changes--conclusions from the

3rd International ESTP Expert Workshop. Toxicologic pathology 40: 971-994

Hayashi S, Fujii E, Kato A, Kimura K, Mizoguchi K, Suzuki M, Sugimoto T, Takanashi H,

Itoh Z, Omura S, Wanibuchi H (2008) Characterization of multinuclear hepatocytes induced

in rats by mitemcinal (GM-611), an erythromycin derivative. Toxicologic pathology 36: 858-

865

HED. (2002) Hepatocellula hypertrophy. HED Guidance Document.

Hissink AM, Kulig BM, Kruse J, Freidig AP, Verwei M, Muijser H, Lammers JH, McKee

RH, Owen DE, Sweeney LM, Salmon F (2009) Physiologically based pharmacokinetic

modeling of cyclohexane as a tool for integrating animal and human test data. International

journal of toxicology 28: 498-509

Holsapple MP, Pitot HC, Cohen SM, Boobis AR, Klaunig JE, Pastoor T, Dellarco VL,

Dragan YP (2006) Mode of action in relevance of rodent liver tumors to human cancer risk.

Toxicological sciences : an official journal of the Society of Toxicology 89: 51-56

Jarnberg, J., Johanson, G., Lof, A., 1996. Toxicokinetics of inhaled trimethylbenzenes in man.

Toxicology and Applied Pharmacology 140: 281-288.

Jarnberg, J., Johanson, G., Lof, A., Stahlbom, B., 1998. Toxicokinetics of 1,2,4-

trimethylbenzene in humans exposed to vapours of white spirit: comparison with exposure to

1,2,4-trimethylbenzene alone. Archives of Toxicology 72: 483-491.

18

Jarnberg, J., Stahlbon, B., Johanson, G., Lof, A., 1997. Urinary excretion of dimethylhippuric

acids in humans after exposure to trimethylbenzenes. International Archives of Occupational

and Environmental Health 69: 491-497.

Johannsen FR, Levinskas GJ (1987) Acute and subchronic toxicity of

tetramethylcyclohexanes. Journal of applied toxicology : JAT 7: 245-248

Johnson W, Jr., Bergfeld WF, Belsito DV, Hill RA, Klaassen CD, Liebler D, Marks JG, Jr.,

Shank RC, Slaga TJ, Snyder PW, Andersen FA (2012) Safety assessment of isoparaffins as

used in cosmetics. International journal of toxicology 31: 269S-295S

Jones, K., Meldrum, M., Baird, E., Cottrell, S., Kaur, P., Plant, N., Dyne, D., Cocker, J., 2006.

Biological Monitoring for Trimethylbenzene Exposure: A Human Volunteer Study and a

Practical Example in the Workplace. Annals of Occupational Hygiene 50: 593-598.

Juran SA, Johanson G, Ernstgard L, Iregren A, van Thriel C (2014) Neurobehavioral

performance in volunteers after inhalation of white spirits with high and low aromatic

content. Archives of toxicology 88: 1127-1140

Kanerva RL, McCracken MS, Alden CL, Stone LC (1987) Morphogenesis of decalin-induced

renal alterations in the male rat. Food and chemical toxicology : an international journal

published for the British Industrial Biological Research Association 25: 53-61

Kennedy GC, Pearce WM, Parrott DM (1958) Liver growth in the lactating rat. J Endocrinol

17: 158-160

Kim J-K, Cho H-W, Han J-H, Lee S-B, Chung Y-H, Rim K-T, Yang J-S (2012) Subchronic

Inhalation Toxicity Study of n-pentane in Rats. Safety and Health at Work 3: 224-234

Kinkead ER, Haun CC, Schneider MG, Vernot EH, MacEwen JD. (1985) Chronic inhalation

exposure of experimental animals to methylcyclohexane. Air Force Aerospace Medical

Research Laboratory, Wright-Patterson Air Force Base, Ohio.

Kostrzewski, P., Wiaderna-Brycht, A., Czerski, B., 1997. Biological monitoring of

experimental human exposure to trimethylbenzene. The Science of the Total Environment

199: 73-81.

Kostrewski, P., Wiaderna-Brycht, A. 1995. Kinetics of elimination of mesitylene and 3,5-

dimethylbenzoic acid after experimental human exposure. Toxicology Letters 77: 259-264.

Kreckman K, Baldwin J, Roberts L, Staab R, Kelly D,Saik J. (2000). Inhalation

developmental toxicity and reproduction studies with cyclohexane. Drug and Chemical

Toxicology 23:555-573.

Lammers J, Emmen H, Muijser H, Hoogendijk E, McKee R, Owen D, Kulig B. (2007).

Neurobehavioral efects of cyclohexane in rat and human. International Journal of Toxicoloy

28:488-497.

19

Lammers JHCM, Muijser H, Owen DE, Kulig BM, McKee RH (2011) Neurobehavioral

Effects of Acute Exposure to Normal (n-) Paraffins. International journal of toxicology 30:

47-58

Lewis RW, Billington R, Debryune E, Gamer A, Lang B, Carpanini F (2002) Recognition of

adverse and nonadverse effects in toxicity studies. Toxicologic pathology 30: 66-74

MacFarland HN, Ulrich CE, Holdsworth CE, Kitchen DN, Halliwell WH, Blum SC (1984) A

Chronic Inhalation Study with Unleaded Gasoline Vapor. International journal of toxicology

3: 231-248

Malley L, Bamberger J, Stadler J, Elliott G, Hansen J, Chiu T, et al. (2000). Subchronic

toxicity of cyclohexane in rats and mice by inhalation exposure. Drug and Chemical

Toxicology 23:513-537.

Maronpot RR, Yoshizawa K, Nyska A, Harada T, Flake G, Mueller G, Singh B, Ward JM

(2010) Hepatic enzyme induction: histopathology. Toxicologic pathology 38: 776-795

McKee R, Frank E, Heath J, Owen D, Przygoda R, Trimmer G, Whitman F (1998).

Toxicology of n-pentane (CAS no. 109-66-0). Journal of Applied Toxicology 18:431-442.

McKee RH, Lammers JH, Muijser H, Owen DE, Kulig BM (2010) Neurobehavioral effects of

acute exposure to aromatic hydrocarbons. International journal of toxicology 29: 277-290

McKee RH, Lammers JH, Muijser H, Owen DE (2011) Neurobehavioral effects of acute

exposure to isoparaffinic and cycloparaffinic hydrocarbons. International journal of

toxicology 30: 715-734

McKee RH, Adenuga MD, Carrillo J-C (2015) Characterization of the toxicological hazards

of hydrocarbon solvents. Critical Reviews in Toxicology 45: 273-365

Melnick RL (1992) Does chemically induced hepatocyte proliferation predict liver

carcinogenesis? FASEB journal : official publication of the Federation of American Societies

for Experimental Biology 6: 2698-2706

Moser GJ, Wolf DC, Harden R, Standeven AM, Mills J, Jirtle RL, Goldsworthy TL (1996a)

Cell proliferation and regulation of negative growth factors in mouse liver foci.

Carcinogenesis 17: 1835-1840

Moser GJ, Wolf DC, Wong BA, Goldsworthy TL (1997) Loss of tumor-promoting activity of

unleaded gasoline in N-nitrosodiethylamine-initiated ovariectomized B6C3F1 mouse liver.

Carcinogenesis 18: 1075-1083

Moser GJ, Wong BA, Wolf DC, Moss WR, Goldsworthy TL (1996b) Comparative short-term

effects of methyl tertiary butyl ether and unleaded gasoline vapor in female B6C3F1 mice.

Fundamental and applied toxicology : official journal of the Society of Toxicology 31: 173-

183

20

Mullin LS, Ader AW, Daughtrey WC, Frost DZ, Greenwood MR (1990) Toxicology update

isoparaffinic hydrocarbons: a summary of physical properties, toxicity studies and human

exposure data. Journal of applied toxicology : JAT 10: 135-142

Nau, C., Neal, J., Thornton, M., 1966. C9-C12 fractions obtained from petroleum distillation.

Archives of Environmental Health 12:382-393.

Nilsen OG, Haugen OA, Zahlsen K, Halgunset J, Helseth A, Aarset H, Eide I (1988) Toxicity

of n-C9 to n-C13 Alkanes in the Rat on Short Term Inhalation. Pharmacology & Toxicology

62: 259-266

NTP (2005) NTP toxicology and carcinogenesis studies of decalin (CAS No. 91-17-8) in

F344/N rats and B6C3F(1) mice and a toxicology study of decalin in male NBR rats

(inhalation studies). National Toxicology Program technical report series: 1-316

Ochs H, Dusterberg B, Schulte-Hermann R (1986) Induction of monooxygenases and growth

in rat liver by progesterone. Archives of toxicology 59: 146-149

Pedersen LM, Cohr KH (1984) Biochemical pattern in experimental exposure of humans to

white spirit. I. The effects of a 6 hours single dose. Acta Pharmacol Toxicol (Copenh) 55:

317-324

Phillips RD, Egan GF (1984a) Effect of C10-C11 isoparaffinic solvent on kidney function in

Fischer 344 rats during eight weeks of inhalation. Toxicol Appl Pharmacol 73: 500-510

Phillips RD, Egan GF (1984b) Subchronic inhalation exposure of dearomatized white spirit

and C10-C11 isoparaffinic hydrocarbon in Sprague-Dawley rats. Fundamental and applied

toxicology : official journal of the Society of Toxicology 4: 808-818

Roberts L, White R, Bui Q, Daughtrey W, Koschier F, Rodney S, Schreiner C, Steup D,

Breglia R, Rhoden R, Schroeder R, Newton P (2001) Developmental toxicity evaluation of

unleaded gasoline vapor in the rat. Reproductive toxicology (Elmsford, NY) 15: 487-494

Russell C. Cattley JAP, Steen L. Vonderfecht (2013) Liver, Gallbladder, and Exocrine

Pancreas. In Toxicologic Pathology: Nonclinical Safety Assessment, Pritam S. Sahota JAP,

Jerry F. Hardisty, Chirukandath Gopinath (ed), 10. Boca Raton, FL: CRC Press

Scampini G, Nava A, Newman AJ, Della Torre P, Mazue G (1993) Multinucleated

hepatocytes induced by rifabutin in rats. Toxicologic pathology 21: 369-376

Schreiner C, Lapadula E, Breglia R, Bui Q, Burnett D, Koschier F, Podhasky P, White R,

Mandella R, Hoffman G (1998) Toxicity evaluation of petroleum blending streams: inhalation

subchronic toxicity/neurotoxicity study of a light alkylate naphtha distillate in rats. Journal of

toxicology and environmental health Part A 55: 277-296

Schulte-Hermann R (1974) Induction of liver growth by xenobiotic compounds and other

stimuli. CRC Crit Rev Toxicol 3: 97-158

21

Schulte-Hermann R (1979) Adaptive liver growth induced by xenobiotic compounds: its

nature and mechanism. Archives of toxicology Supplement = Archiv fur Toxikologie

Supplement: 113-124

Standeven AM, Blazer DG, 3rd, Goldsworthy TL (1994a) Investigation of antiestrogenic

properties of unleaded gasoline in female mice. Toxicol Appl Pharmacol 127: 233-240

Standeven AM, Goldsworthy TL (1993) Promotion of preneoplastic lesions and induction of

CYP2B by unleaded gasoline vapor in female B6C3F1 mouse liver. Carcinogenesis 14:

2137-2141

Standeven AM, Goldsworthy TL (1994) Identification of hepatic mitogenic and cytochrome

P‐450‐inducing fractions of unleaded gasoline in B6C3F1 mice. Journal of toxicology and

environmental health 43: 213-224

Standeven AM, Wolf DC, Goldsworthy TL (1994b) Interactive effects of unleaded gasoline

and estrogen on liver tumor promotion in female B6C3F1 mice. Cancer research 54: 1198-

1204

Standeven AM, Wolf DC, Goldsworthy TL (1995) Promotion of hepatic preneoplastic lesions

in male B6C3F1 mice by unleaded gasoline. Environmental health perspectives 103: 696-700

Stone LC, Kanerva RL, Burns JL, Alden CL (1987a) Decalin-induced nephrotoxicity: light

and electron microscopic examination of the effects of oral dosing on the development of

kidney lesions in the rat. Food and chemical toxicology : an international journal published

for the British Industrial Biological Research Association 25: 43-52

Stone LC, McCracken MS, Kanerva RL, Alden CL (1987b) Development of a short-term

model of decalin inhalation nephrotoxicity in the male rat. Food and chemical toxicology : an

international journal published for the British Industrial Biological Research Association 25:

35-41

Stuchal LD, Weil RE, Roberts SM (2013) Derivation of a chronic reference concentration for

decalin. Regulatory toxicology and pharmacology : RTP 66: 38-46

Sung JH, Choi B-G, Kim HY, Baek M-W, Ryu HY, Kim YS, Choi YK, Yu IJ, Song KS

(2010) Acute and Subchronic Inhalation Toxicity of n-Octane in Rats. Safety and Health at

Work 1: 192-200

Swenberg JA, Lehman-McKeeman LD (1999) alpha 2-Urinary globulin-associated

nephropathy as a mechanism of renal tubule cell carcinogenesis in male rats. IARC Sci Publ:

95-118

Thoolen B, Maronpot RR, Harada T, Nyska A, Rousseaux C, Nolte T, Malarkey DE,

Kaufmann W, Kuttler K, Deschl U, Nakae D, Gregson R, Vinlove MP, Brix AE, Singh B,

Belpoggi F, Ward JM (2010) Proliferative and nonproliferative lesions of the rat and mouse

hepatobiliary system. Toxicologic pathology 38: 5S-81S

22

Tilbury L, Butterworth BE, Moss O, Goldsworthy TL (1993) Hepatocyte cell proliferation in

mice after inhalation exposure to unleaded gasoline vapor. Journal of toxicology and

environmental health 38: 293-307

Treon JF, Crutchfiend WE, Kitzmiller KV (1943) The physiological response of animals to

cyclohexane, methylcyclohexane, and certain derivatives of these compounds. II. Inhalation J

Ind Hyg Toxicol 25: 323-347

UKHSE. (2005) Toxicity Working Documents. UK Guidance Documents: PSD guidance

document: Interpretation of liver enlargement in regulatory toxicity studies.

USEPA. (1991) Alpha2u-globulin: Association with chemically induced renal toxicity and

neoplasia in the male rat. Risk Assessment Forum, USEPA, Washington, DC.

USEPA. (2007) Toxicological review of 2,2,4-trimethylpentane (CAS No. 540-84-1). U.S.

Environmental Protection Agency, Washington, DC.

White AC, Gershbein LL (1987) Steroid modulation of liver regeneration and hepatic

microsomal enzymes in rats of either sex. Res Commun Chem Pathol Pharmacol 55: 317-334

Williams GM, Iatropoulos MJ (2002) Alteration of liver cell function and proliferation:

differentiation between adaptation and toxicity. Toxicologic pathology 30: 41-53

23

Appendix I

Trimethylpentane

Background

No MAK or TRGS 900 values are available for trimethylpentanes (isooctane). While no

explanation for this is provided in TRGS 900, all trimethylpentane isomers are classified 3A5

for carcinogenicity. In the DFG’s justification for the cancer classification of

trimethylpentane, the critical effect was determined to be liver carcinogenicity observed in

female but not male mice exposed to 0, 67, 292 or 2056 ml/m3 (approximately 0, 273, 1190 or

8400 mg/m3) unleaded gasoline containing four isomers (2,3,4-, 2,2,4-, 2,3,3- and 2,2,3-

isomers accounting for 12% of unleaded gasoline by volume) of trimethylpentane.

Liver tumors in female mice as a direct consequence of hepatocellular proliferation following

exposure to unleaded gasoline

Although there are no chronic studies of specific isomers of trimethylpentanes, the DFG has

concluded that the liver tumorigenicity of unleaded gasoline was selectively caused by the

hepatocellular proliferative activity of trimethylpentanes on the basis of a short term tumor

promotion study in the liver of female mice (DFG, 2002). This hypothesis is also based on the

absence of genotoxic activity for unleaded gasoline and 2,2,4-trimethylpentane (DFG, 2002).

To test this hypothesis, Standeven and Goldsworthy conducted a series of studies that showed

that unleaded gasoline (UG), particularly boiling fractions containing trimethylpentanes,

could induce an increase in the incidence of gross pathological lesions of the liver when

exposure was preceded by treatment with a tumor initiator, compared to mice treated with the

initiator alone (Standeven & Goldsworthy, 1993; Standeven & Goldsworthy, 1994). The

induction of these lesions also appeared to be associated with relative liver weight increase,

increased hepatocellular proliferation, several-fold increase in pentoxyresorufin-O-dealkylase

(PROD) activity, and increased cytochrome P450 content in the liver.

However, there are several problems with the hepatocellular proliferation hypothesis. If it is

true that the UG-induced liver tumors in female mice are a direct consequence of

hepatocellular proliferation as hypothesized by Standeven and Goldsworth, it would be

expected that no such proliferation should occur in the liver of male mice since UG was not

identified as a hepatocarcinogen in males. In fact, when Standeven et al. (1995) tested this

theory by repeating the same study described in Standeven and Goldsworthy (1993) in male

mice, practically identical results were found. Numbers of gross pathological lesions were

statistically significantly increased and the relative liver weights, hepatocyte proliferation and

PROD activity in microsomal liver fractions were all increased. Tilbury et al. (1993)

conducted a 13-week subchronic inhalation toxicity study in male and female mice exposed to

identical concentrations of UG. As previously reported, hepatocyte proliferation and liver

weight increases were noted in both males and females with no discernable sex differences. It

should be noted that none of the studies with UG alone reported any evidence for

5 Category 3 substances are those that cause concern that they could be carcinogenic for man but

cannot be assess conclusively because of lack of data.

3A – Substances that cause cancer in humans or animals or that are considered to be carcinogenic for

humans for which the criteria for classification in category 4 or 5 are in principle fulfilled. However,

the database for these substances is insufficient for the establishment of a MAK or BAT value.

24

hepatotoxicity in the form of increased serum levels of liver enzymes or significant

histopathological changes in the liver. Even in the 2-year carcinogenicity study, no UG-

related effects were noted in any organ in male mice, indicating that with the exception of the

liver tumors in female mice, UG vapor up to 8400 mg/m3 is practically nontoxic to rats and

male mice, discounting the kidney effects in male rats that are not toxicologically relevant

(MacFarland et al, 1984).

In discussing whether hepatocyte proliferation is indeed a predictor for liver carcinogenesis,

Melnick (1992) describes the two types of proliferative stimuli that occur in the liver: the first

involves regenerative or compensatory growth such as would occur following loss of

hepatocytes due to overt toxicity or partial hepatectomy while the second type involves a

direct mitogenic hyperplasia, leading to liver enlargement without any evidence for tissue

toxicity. Furthermore, Melnick, (1992) considered the latter as a normal stimulation of liver

growth (hypertrophy and/or hyperplasia) and a functional adaptation to an increase in

metabolic load following exposure to a xenobiotic. Considering the lack of evidence for overt

liver toxicity in male mice in the subchronic and chronic toxicity studies of UG by Tillbury et

al. (1993) and MacFarland et al. (1984), we can conclude that only the second type of

proliferative stimuli described by Melnick (1992) occurs in mice following UG vapor

exposure. The idea postulated by Melnick is not new, in fact; Schulte-Hermann (1974)

described this exact form of liver growth, which is not associated with pathological changes

in the liver, as a form of “additive” liver growth in response to increased metabolic load

(Schulte-Hermann, 1974; Schulte-Hermann, 1979). In reviewing several chemical-induced

histopathological liver lesions, Williams & Iatropoulos, (2002) described the characteristics

of a typical adaptive liver response to include increased and reversible liver growth to

enhance the capacity to respond to stress which does not otherwise compromise liver

function. In their analysis of liver enlargement due to both hypertrophy and hyperplasia, the

authors speculated that the need to adapt to a higher metabolic load induces the hyperplasia of

organelles, particularly the endoplasmic reticulum and peroxisomes. Interestingly, this form

of mitogenic hyperplasia in the liver does not occur only in the event of chemical exposure.

Argyris, (1971) described a study where two sets of 4-week old rats were maintained on a low

(15%) and high (64%) protein diet for 7 days. Rats maintained on the high protein diet

showed a 30% increase in absolute and relative liver weights, which was associated with an

increase in mitotic activity and induction of enzymes involved in the urea cycle. Kennedy et

al, (1958) observed a similar increase in hypertrophy and mitotic activity in hepatocytes

during the second and third weeks of lactation in rat pups. The authors concluded that these

effects were most likely related to the changes in food intake.

In general, there is sufficient evidence to indicate that the formation of altered hepatic foci

and hepatocellular proliferation is not indicative of liver carcinogenesis in female mice with

long term exposure in the case of UG exposure. With regard to altered hepatic foci, if as

Standeven and others have postulated that altered hepatic foci represent likely preneoplastic

lesions, an alternative mechanism must explain why no such tumors are found in male mice

even though the initiation-promotion studies showed a stronger response (in terms of altered

hepatic foci) in male mice compared to female mice. With regard to hepatocellular

proliferation, one of the important characteristics that must be met in establishing a causal

relation between a set of key events and adverse outcome in a mode of action is biological

gradient. In other words, one would expect that induction of liver tumors would be preceded

by a sustained period of dose-responsive hepatocellular proliferation if indeed there was a

25

causal relationship. However, as indicated by Tilbury et al. (1993), early evidence for

hepatocyte proliferation (after one week of exposure) was not sustained with prolonged

exposure.

Estrogen as a potential explanation for the lack of liver tumors in male mice

Following evidence showing no difference in the hepatocyte proliferative activity in the livers

of male and female mice that could account for the female-specificity of the liver tumors,

Standeven, Goldsworthy and others proposed that the lack of liver tumors in male mice may

have been due to differences in sensitivities of the initiation-promotion studies and other

factors such as estrogen in female mice (Moser et al, 1996b; Standeven et al, 1995). In fact,

the sex hormone hypothesis is the basis for the DFG’s classification of trimethylpentanes as a

tumor promoter. Two types of studies were conducted to evaluate the estrogen hypothesis: in

the first set of studies, isolated hepatocytes from female mice administered 1800 mg/kg/day

of UG for three days showed a 3-fold increase in the rate of metabolism of 17-β estradiol

compared to rats administered corn oil (Moser et al, 1996b; Standeven et al, 1994a). This was

not surprising considering that UG, similar to other hydrocarbons, is an inducer of liver

metabolizing enzymes. Oral UG treatment in mice had no functional anti-estrogenic effects in

in vivo uterotrophic assays. The authors concluded that if UG causes uterine effects in female

mice; it is not likely related to anti-estrogenic effects (Standeven et al, 1994a). The second set

of studies was modeled after the initiation-promotion studies described earlier by Standeven

and others. However, in this case, a separate set of female mice were exposed to 1 mg/kg

doses of estradiol in the diet during the 16 week inhalation exposure to 8400 mg/m3 UG

preceded by intraperitoneal exposure to DEN initiator. The authors concluded that estradiol

co-exposure with UG strongly potentiated liver tumor promotion in female mice (Moser et al,

1996a; Standeven et al, 1994b). To support this conclusion, the same study was repeated in

ovariectomized mice (to model lower estrogen levels). The authors concluded that the

significant attenuation of the volume fraction of liver occupied by hepatic foci (as a measure

of the tumor promoting activity of UG) in ovariectomized mice compared to intact mice, was

evidence that the tumor promoting activity of UG was dependent on its interaction with sex

hormones specific to female mice (Moser et al, 1997). However, it is not clear how this

conclusion is reached considering there does not appear to be any statistically significant

difference between intact initiated mice exposed to UG, initiated ovariectomized mice and

initiated ovariectomized mice exposed to UG for any of the parameters measured.

In general, the in vivo estradiol studies appear to be quite unclear and inconsistent. In Moser

et al. (1997), the authors indicate that the mechanism of UG tumor promoting activity is likely

due to its anti-estrogenic effect. However, other than the increased metabolism of estradiol in

isolated hepatocytes of mice exposed to UG, no in vivo evidence is provided. In fact, Moser et

al. (1997) found no effects of UG exposure on several possible parameters that may indicate

anti-estrogenic effects mediated through the estrogen receptor, simply concluding that a

mechanism for estrogen antagonism is unknown. In Standeven et al. (1994b), the authors

showed that co-exposure with estradiol (1 mg/kg) massively potentiated the liver tumor

promoting activity of UG in initiated mice (6-fold higher mean volume of foci compared to

UG alone). However, in the same study, treating initiated mice with estradiol alone

significantly attenuated mean volume of foci by 4-fold compared to initiated mice alone. If, as

has been speculated by the authors in the in vitro studies (Standeven et al, 1994a) and the

DFG that the key tumor promoting effect of UG is to increase the metabolism of estrogen,

one would expect that co-exposure of UG and estrogen in initiated mice should significantly

26

decrease tumor promotion compared to exposure of initiated mice to UG alone since the

higher rate of estrogen metabolism should be compensated by the administration of

exogenous estrogen. Furthermore, there were no differences in the tumor-promoting

parameters measured in the studies of male and female mice. If increased estrogen

metabolism is key to the tumor promoting activities of UG, what is the mechanism by which

remarkably similar lesions are seen in male mice in identical initiation promotion studies?

It appeared that over time, the Goldsworthy laboratory (which published all the key papers

cited by the DFG in its opinion on trimethylpentanes) changed its opinion on several initial

claims. For example, Moser and others categorically stated in their paper that they no longer

considered the UG-dependent increase in microsomal liver enzyme activity, liver weight

increase and increase in hepatocellular proliferation (as measured by labeling indices) as

specific markers of the tumor promoting activity of UG (Moser et al, 1997). This was due in

part by the fact that these effects were observed in ovariectomized mice to the same degree as

observed in intact mice. The fact that these effects are also seen in male mice, even though

they do not develop liver tumors may have also been a likely factor although unmentioned. In

Standeven and Goldsworthy (1994), the authors had indicated that specific boiling point

fractions containing trimethylpentanes (100 °C < BP < 132 °C) were responsible for the

tumor promoting effects of UG. However, Moser et al. (1997) had utilized a different

formulation of UG that was composed of slightly higher percentages of aromatics and olefins

and a lower percentage of saturated hydrocarbons. Although the composition of this particular

UG was not given, it appeared that the virtually identical results seen with this formulation

compared to the earlier formulations that contained 12% trimethylpentane isomers in the

liquid form must have indicated to the authors that perhaps trimethylpentanes were not likely

the specific component responsible for the effects observed in mice following UG exposure.

The authors concluded thus “however, the specific component or components of UG which

are responsible for the hepatic tumor-promoting ability of UG are unknown”. This point is

very important to explain in more detail. In Standeven and Goldsworthy (1994), the UG

formulation used was PS-6 which according to the analysis conducted by the American

Petroleum Institute (API) in 1993 was composed of 14 wt.% trimethylpentane isomers

(includes trace levels of C8 olefins) (API, 2008). For the formulation used in Moser et al.

(1997), two API analyses are available. In the 1991 analysis (API 1991-01A),

trimethylpentane isomers constituted approximately 1.5% of total liquid by weight. In the

1994 analysis (API 1991-01B) however, trimethylpentane isomer composition was

approximately 4.1% of total liquid by weight (API, 2008). Assuming API 1991-01B as the

formulation tested in Moser et al. (1997), it is noteworthy that the authors themselves

acknowledge that virtually identical results were obtained as when PS-6 was tested despite the

clear 3.5-fold difference in trimethylpentane content. More importantly, while the Standeven

and Goldsworthy (1994) study that identified the fraction responsible for UG’s tumor

promoting activity was conducted via intragastric intubation (I.e. oral exposure), all the

initiation promotion studies in intact male/female and ovariectomized mice were conducted

via vapor inhalation. This is a critical point as available compositional data on several UG

formulations has shown that the liquid composition of UG formulations is not identical to the

vapor composition. As an example, Roberts et al, (2001) provides a comparison between the

liquid and vapor composition of another UG formulation (API 94-02). While the C8 fraction

(including saturated and unsaturated hydrocarbons) comprises 20.7% of the liquid

formulation, it only makes up 3.1% of the vapor phase. On the other hand, the C5 fraction

which makes up 16.8% of the liquid formulation, represents >46% of the vapor phase. In

27

other words, the volatile fraction of UG that would have been responsible for the effects seen

in the initiation promotion studies is primarily made up of volatile short chain C4-C6

aliphatics (approximately 89% in Roberts et al. 2001).

Key conclusions

[1] As stated by the authors (Moser et al. 1997), it is clear that hepatocellular

proliferation, with regard to UG exposure in mice, is not a marker for carcinogenesis

following long term exposure to unleaded gasoline. Not only is this liver lesion

(which have not been shown in any studies so far to be associated with

histopathological lesions or increased serum levels of liver enzymes consistent with

overt liver toxicity) well described in published literature as indicative of an adaptive

change to accommodate increased metabolic load, there is no current explanation for

why such lesions occur in male mice which do not develop liver tumors following

prolonged exposure. The adaptive effects of UG on the liver are not specific to

gasoline alone. Several other studies have shown that hydrocarbons, including

aromatics, alcohols (which are strong inducers of Cyp 2E1) such as ethanol and

aliphatics induce an increase in liver metabolism following either oral or inhalation

exposure. These effects have generally been regarded as functional responses and are

not indicative of adverse effects in the liver, most especially as these mitogenic

effects are also seen with physiological stimuli such as increased protein intake.

[2] With regard to trimethylpentanes specifically, the USEPA recently completed a

comprehensive review of the studies on 224-trimethylpentane, where some of the

Standeven studies were reviewed (USEPA, 2007). The overall conclusion was that

the mitogenic effects of 224-trimethylpentane was not indicative of an adverse liver

effect and the male rat-specific kidney effect was considered the critical health effects

of trimethylpentanes.

There were no differences in the results obtained in the initiation promotion studies despite

the distinct difference in the levels of trimethylpentanes contained in the PS-6 and 91-1 UG

formulations. As stated by the authors, this indicates that trimethylpentanes are not likely to

be the specific components responsible for the tumor promoting effects ascribed to UG. In

addition, the vapor phase of UG mostly contains highly volatile C4-C6 aliphatics, providing

stronger evidence that the suggested tumor promoting effects ascribed to UG are likely not

related to trimethylpentane content.