(98-023 / 6-30-99)

Risk Assessment Issue Paper for:
Derivation of a Provisional RfD for
1,2,4-Trimethylbenzene (CASRN 95-63-6) and 1,3,5-Trimethylbenzene (CASRN 108-67-8)

NOTE: This provisional risk assessment paper for the oral chronic reference dose (RfD) for 1,2,4-Trimethylbenzene and 1,3,5-Trimethylbenzene has received external peer review.

INTRODUCTION

An RfC for 1,2,4-trimethylbenzene (TMB), 1,3,5-TMB, or a mixture of the TMBs is unavailable on IRIS (U.S. EPA, 1998) or in HEAST (U.S. EPA, 1997). EPA's CARA database (U.S. EPA, 1994) lists a HEA on TMBs and a drinking water health advisory document for 1,3,5-TMB and 1,2,4-TMB (U.S. EPA, 1987a; 1987b). None of these compounds have been the subject of an ATSDR toxicological profile. ACGIH (1998) has adopted a TLV-TWA of 25 ppm (123 mg/m³) for mixed isomers of TMB. NIOSH has established a REL-TWA of 25 ppm (123 mg/m³) for both 1,2,4-TMB and 1,3,5-TMB (NIOSH, 1994). OSHA (1997) has not established a PEL for either 1,2,4-TMB or 1,3,5-TMB.

Research papers pertinent to the derivation of a provisional RfC for 1,2,4-TMB and 1,3,5-TMB were sought through computer searches of the HSDB, RTECS, TSCATS, MEDLINE, and TOXLINE (and its subfiles) databases covering the time period of 1992-1998. The literature search was conducted in October 1998.

REVIEW OF PERTINENT INFORMATION

Production and use of 1,2,4-TMB as an intermediate in the manufacture of trimellitic anhydride, dyes, pharmaceuticals, and pseudocumidine, and its chief industrial use as a solvent and paint thinner may result in its release to the environment through various waste streams. Release of 1,2,4-TMB to the environment also occurs due to its use as a component of gasoline and as emissions from gasoline-powered vehicles, municipal waste-treatment plants, and coal-fired power stations (HSDB, 1998).

TOXICOKINETICS

Järnberg et al. (1996) studied the fate of 1,2,4-TMB, 1,3,5-TMB and 1,2,3-TMB in humans during and after exposure. Ten healthy male volunteers were exposed on separate occasions to either 1,2,4-TMB (>99% purity), 1,3,5-TMB (99% purity), or 1,2,3-TMB (90-95% purity) at a concentration of 25 ppm (123 mg/m³) or to 1,2,4-TMB at a concentration of 2 ppm (9.8 mg/m³). Exposures were for 2 hours in a 20 m³ chamber at a constant work load of 50 W on an ergometer bicycle. The work load was intended to simulate light work. Respiratory uptake, the sum of net uptake plus amount cleared by exhalation during exposure, was 64%, 62% and 56% for 1,2,4-TMB, 1,3,5-TMB and 1,2,3-TMB respectively. The concentration of TMB in capillary blood increased rapidly at the beginning of the exposure, then leveled off without reaching a plateau within the 2-hour exposure. The elimination process was separated into four compartments reflecting washout from lung alveoli as well as distribution from rapidly perfused, slowly perfused and adipose tissues. Half-lives for elimination from the four compartments were calculated to be: 1-2 min, 21-22 min, 4-5 hours, and a terminal half-life of 78-120 hours. No significant differences in uptake or elimination half-lives between the different exposure concentrations of 1,2,4-TMB were observed.

Järnberg et al. (1996) also evaluated irritation and central nervous system (CNS) effects by administering a questionnaire before, during and after exposure. The questionnaire covered discomfort in eyes, nose, throat and airways, as well as questions concerning headache, fatigue, nausea, dizziness, intoxication, difficulty in breathing and smell of the solvent. No irritation or CNS effects were reported by the subjects under these exposure conditions.

In a subsequent paper, Järnberg et al. (1997) reported urinary excretion of dimethylhippuric acids (DMHAs) in the humans that had been exposed to TMB. All urine was collected from initiation of exposure until 20 hours postexposure for 25 ppm or until 4 hours postexposure following exposure at 2 ppm 1,2,4-TMB. Urine was analyzed for excretion of unconjugated dimethylbenzoic acids (DMBAs) and for DMHAs by HPLC. The total respiratory uptake or dose was calculated as the sum of net uptake and the amount cleared by exhalation during exposure. Half-times were calculated from the slope of the semilogarithmic excretion-time curve. ANOVA was used to compare mean values recorded for the two exposure levels of 1,2,4-TMB. Student's paired t-test was used to compare means for the half-times for excretion, the level of significance being set to p<0.05.

A major metabolic pathway for the TMBs is aliphatic hydroxylation of one of the methyl groups followed by further oxidation to DMBA. The DMBAs may undergo conjugation with glycine to yield DMHAs. Excretion of the DMHAs has been suggested as biomarkers for TMB exposure and the DMHAs are not endogenously formed. On the average, between 3 and 22% of the respiratory uptake of the three TMBs was excreted as DMHAs within 24 hours. The main DMHA excreted following 1,2,4-TMB exposure was 3,4-DMHA, accounting for 18% of the dose. The main DMHA isomer excreted following 1,2,3-TMB exposure was 2,3-DMHA accounting for 9% of the absorbed dose. Excretion of 3,5-DMHA, the only possible isomer, following 1,3,5-TMB exposure accounted for only 3% of the absorbed dose. The half-times of excretion rate were in the range of 4-8 hours for all DMHA isomers formed from 1,2,3- and 1,2,4-TMB. In contrast, 3,5-DMHA had an average half-time of 16 hours. The cumulative excretion pattern of DMHA isomers did not differ significantly between the 25 and 2 ppm exposure levels of 1,2,4-TMB, which suggests that the kinetics are linear with dose. The results suggest that urinary excretion of DMHA in humans may serve as a good indicator of recent exposure to TMB although the sum of the DMHA isomers might be preferred to single isomers.

Kostrewski et al. (1997) investigated the toxicokinetics of 1,2,4-TMB, 1,3,5-TMB, and 1,2,3-TMB with human volunteers (5/group) exposed to concentrations of 5-150 mg/ m³ for 8 hours. Exhaled air, capillary blood and urine samples were collected before, during and after the exposures. The retention of 1,2,4-TMB, 1,3,5-TMB and 1,2,3-TMB in the lungs was 68%, 67% and 71%, respectively. Elimination of TMB from capillary blood followed a pattern fitting an open three-component model. Elimination from the three compartments occurred with half-lives of: 0.016, 0.33, and 44.1 hours for 1,2,4-TMB; 0.03, 0.72, and 46.2 hours for 1,3,5-TMB; and 0.04, 1.7, and 28.9 hours for 1,2,3-TMB. Urinary excretion of the TMB isomers follows oxidation to dimethyl benzene alcohols or DMBAs, and conjugation with glucuronic acid or sulfate. Urinary excretion of DMBA was described by a two-compartment model with half-lives ranging from 2.2-6.5 hours for the first compartment and ranging from 34-63 hours for the second compartment. Based on the kinetic data, they also developed a 14-day simulation model of accretion and excretion of DMBA in urine. The model predicted that the highest rates of DMBA excretion would occur on day 5 of exposure.

Human Toxicity Data

There is a paucity of information available on the toxicity of 1,2,4-TMB and 1,3,5-TMB in humans and animals. Human data are limited to an occupational exposure study in which workers were exposed to a mixture of TMB isomers (Bättig et al., 1958), and no data on the oral toxicity of TMB in humans were located.

Bättig et al. (1958) examined 27 workers exposed to Fleet-X DV 99 solvent in the painting shop of a Swiss transportation plant. The solvent was analyzed spectrographically and was found to consist primarily of aromatic hydrocarbons (97.5%) and paraffinic and naphthenic hydrocarbons (2.5%). The aromatic hydrocarbon portion was comprised of 1,2,4-TMB (>50%), 1,3,5-TMB (>30%), 1,2,3-TMB, 1-methyl-2-ethyl benzene, 1-methyl-3-ethyl benzene, and 1-methyl-4-ethyl benzene. Based on analysis of air samples collected from the plant, the concentration of the solvent was 10-60 ppm (49-295 mg/m³). The control group consisted of 10 unskilled workers employed in a different section of the plant. Although the authors stated that the Fleet-X DV 99 solvent was used for "a period of some ten years", the average exposure duration of the workers was not reported. The workers reported CNS symptoms (vertigo, headaches, and drowsiness) more often than the control group (70% versus 30% in the controls). Chronic asthma-like bronchitis (30% of workers versus 10% of controls), anemia with < 4.5 million erythrocytes/mm³ usually combined with normal hemoglobin (50% versus 20%), and alterations in blood clotting (30% versus 10%) were also observed in the exposed workers. The incidence of CNS symptoms was statistically significantly higher in the exposed workers than in the control group (Fisher's exact test; p<0.05). For the other effects, the incidences did not significantly differ between the groups. A higher incidence of vitamin C deficiency was observed in the control group, suggesting that the two groups may not have been matched for socioeconomic status. If the assumption is made that the solvent exclusively contained TMB isomers, then this study identifies a LOAEL of 10 ppm for signs of neurotoxicity.

Laboratory Animal Toxicity Data

1,3,5-TMB Two studies in which animals received repeated oral doses of 1,3,5-trimethylbenzene were located. Changes in microsomal enzyme activity was the primary focus of one of these studies (Pyykkö, 1980); the second study is an oral subchronic study which examined a number of relevant endpoints (IITRI, 1995). No oral or inhalation developmental or reproductive toxicity studies were located for 1,3,5-trimethylbenzene or other trimethylbenzene isomers.

Pyykkö (1980) administered 0 or 10 mmol/kg/day (0 or 1202 mg/kg-day) 1,3,5-trimethylbenzene in corn oil to 8-10 rats for 3 days. No effects on survival or behavior were noted in the treated or control groups. The study design involved fasting periods which resulted in weight loss in both groups; the weight loss was significantly higher in the 1,3,5-trimethylbenzene group than in the control group. Treatment with 1,3,5-trimethylbenzene also resulted in significant increases in liver weight and in the cytochrome b₅ content of the liver and kidney. Increases in the activity of various microsomal enzymes in the liver, kidneys and lungs were also reported in the animals dosed with 1,3,5-trimethylbenzene.

In the subchronic study on the oral toxicity of 1,3,5-TMB conducted by IIT Research Institute (IITRI, 1995), groups of 10 male and 10 female Sprague Dawley rats were administered via gavage 0, 50, 200, or 600 mg/kg 1,3,5-TMB in corn oil 5 days/week for 90 days. An additional group of rats (10/sex) were administered 600 mg/kg 1,3,5-TMB for 90 days and retained without treatment for 28 days. Duration-adjusted doses were 0, 36, 143, and 429 mg/kg-day. The following parameters were used to assess toxicity: physical examinations and clinical observations, ophthalmological examinations, body weights, food consumption, hematological and clinical chemistry, organ weights, and gross and histopathology. No compound-related deaths were observed, and abnormal clinical observations consisted of discolored and/or wet inguinal fur and salivation observed in the 600 mg/kg group. A significant decrease in body weight gain (approximately 11% lower body weight gain as compared to controls) was observed in the 600 mg/kg group. In the male rats dosed with 600 mg/kg, significant increases in serum alkaline phosphatase and phosphorus levels and a decrease in glucose level were observed. In the female rats dosed with 600 mg/kg, significant increases in serum cholesterol and phosphorus levels and decreases in sodium and chloride levels were observed. No significant alterations in serum chemistry parameters were observed at the lower dosage levels. The authors noted that with the exception of the increased serum phosphorus levels, the altered serum chemistry values were within normal historical ranges or were due to very high values for two individual animals. The increased phosphorus levels were considered to be treatment-related; the serum phosphorus levels were 17 and 23% higher in the 600 mg/kg male and female rats, respectively, than in the control group. No significant alterations in erythrocyte parameters or total leukocyte levels were observed. A significant increase in monocyte levels were observed in the 200 and 600 mg/kg groups. The authors noted that this increase was not considered treatment-related; however, no explanation was given for this conclusion. No significant alterations in hematological or serum chemistry parameters were observed in the 600 mg/kg recovery group. In the 600 mg/kg group, significant increases in absolute (females only) and relative liver weights and relative kidney weights (males only) were observed. No significant alterations in organ weights were observed in the 600 mg/kg recovery group. No 1,3,5-TMB related alterations in gross or histopathology were observed.

The toxicological significance of the increased serum phosphorus is not known. Although increased kidney weights were observed, no biologically significant increases in other serum electrolytes were observed, suggesting that the increased serum phosphorus levels may not have been related to renal damage. Damage to a number of other tissues, such as the bone, parathyroid or thyroid, could have caused the increased serum phosphorus levels. The finding that the serum phosphorus levels returned to normal in the 600 mg/kg recovery group suggests that the serum phosphorus level alterations were related to 1,3,5-TMB exposure. Taken together, the slight decrease in body weight gain, clinical observations, increased serum phosphorus levels, and increased liver and kidney weight, suggest that the 600 mg/kg (429 mg/kg-day) dose is a LOAEL for 1,3,5-TMB. The NOAEL is 200 mg/kg (143 mg/kg-day).

1,2,4-TMB The database of repeated oral exposure studies in animals for 1,2,4-TMB is limited to an oral exposure study (Borriston Laboratories, Inc., 1984) and a chronic exposure carcinogenicity study (Maltoni et al., 1997). In addition, the oral toxicity of 1,3,5-TMB has been examined in a subchronic study (IITRI, 1995). No oral or inhalation developmental or reproductive toxicity studies were located for 1,2,4-TMB.

The primary focus of the Borriston Laboratories study (1984) was the ability of 1,2,4-TMB to induce nephrotoxicity. In this study, groups of 10 male Fischer-344 rats were administered doses of 0.5 or 2.0 g/kg neat 1,2,4-TMB by gavage 5 days/week for 4 weeks; the duration-adjusted doses were 357 and 1429 mg/kg-day, respectively. A group of rats serving as controls were gavaged with saline. Gross necropsy was conducted in all rats, but only the kidneys underwent histopathological examination. Mortality rates during treatment in the control, low-, and high-dose groups were 0/10, 1/10, and 10/10, respectively. Deaths in the high-dose group occurred as early as the third day of treatment. Final body weight and absolute kidney weight of low-dose rats were not significantly different than controls. Gross necropsy findings in low-dose animals included speckled cortical surfaces in the kidneys and white gelatinous material inside the urinary bladders. High-dose rats exhibited mottled and red thymus, spotty kidney and liver surfaces, enlarged adrenals, gas filled and yellow intestines, and lung congestion. The presence or absence of hydrocarbon nephropathy was determined by examining the incidence of hyaline droplet changes, regenerative epithelium, and tubular dilation with granular material. Treatment with 1,2,4-TMB did not significantly increase the incidence or severity of nephropathy relative to controls; however, according to the authors, it is possible that high-dose rats died before nephropathy could develop.

Maltoni et al. (1997) investigated the carcinogenicity of 1,2,4-TMB (99% pure) in a long-term oral exposure experiment. Male and female Sprague-Dawley rats (50/sex/group) received doses of either 0 or 800 mg/kg (4 days/week) of 1,2,4-TMB by gavage in 1 ml olive oil for 104 weeks. Food and water consumption and body weights were recorded throughout the experiment; however, no data were presented in the report. Animals were terminated after 123 weeks. Upon death the animals were subjected to systemic necropsy. Histopathology was routinely performed on brain, pituitary gland, Zymbal glands, salivary glands, Harderian glands, head, tongue, thymus, mediastinal lymph nodes, lung, heart, diaphragm, liver, spleen, pancreas, kidneys, adrenal glands, esophagus, stomach, intestine (four levels) bladder, prostrate, uterus, vagina, gonads, interscapular fat pad, subcutaneous and mesenteric lymph nodes, sternum, femur, spinal cord, and any other organs and tissues with pathological lesions. No statistical analysis of the data was presented.

Gavage exposure to 1,2,4-TMB (800 mg/kg) for 104 weeks resulted in a slight reduction in the survival of the female Sprague-Dawley rats and an intermediate reduction in the survival of male rats. Quantitative survival data were not presented in the report, and no statistical analysis of the decreases in survival were presented. However, due to the seriousness of decreased survival as a toxicological endpoint, the exposure level of 800 mg/kg is considered to be a LOAEL. The data presented and the fact that only one dose level was used make this study unacceptable as the basis for deriving an RfD.

DERIVATION OF PROVISIONAL RfDs

1,3,5-TMB There is limited information on the oral toxicity of 1,3,5-trimethylbenzene in humans and animals. Only one long-term oral study was identified. This study (IITRI, 1995) identified a NOAEL of 200 mg/kg (143 mg/kg-day) and a LOAEL of 600 mg/kg (429 mg/kg-day) for a slight decrease in body weight gain, clinical observations (discolored/wet inguinal fur and salivation), increased liver and kidney weights, and increased serum phosphorus levels. These effects were not observed in a second group of rats dosed with 600 mg/kg for 90 days and then allowed to recover for 28 days.

A provisional RfD can be derived by dividing the NOAEL of 143 mg/kg-day (IITRI, 1995) by an uncertainty factor of 3000 (10 to account for extrapolation from a subchronic study, 10 for interspecies extrapolation, 10 for human variability, and 3 for database deficiencies); the resultant provisional RfD would be 5E-2 mg/kg-day. The limitations of the database include the lack of a long-term general toxicology study in a second species, developmental toxicity studies in two species, and a multigeneration reproductive toxicity study. Confidence in the principal study is high; it is a well-designed study examining a number of relevant endpoints. Confidence in the database is low; only one long-term oral study was located and no information is available on the potential of ingested 1,3,5-trimethylbenzene to induce developmental, reproductive, or neurological effects (a sensitive endpoint in inhalation studies). Reflecting the low confidence in the database, confidence in the RfD is low.

1,2,4-TMB There is limited information on the oral toxicity of 1,2,4-TMB in humans and animals. The nephrotoxicity study by Borriston Laboratories (1984) is too limited in scope to be used to identify a NOAEL or LOAEL for 1,2,4-TMB; although the 2000 mg/kg dose (1430 mg/kg-day) is clearly a fatal effects level (FEL) for increased mortality. The study of Maltoni et al. (1997) is also unsuitable for derivation of an RfD, as only one dose level was employed. Thus, the database for 1,2,4-TMB is inadequate to derive a provisional RfD using oral exposure data. However, two other approaches could be used to derive an RfD. The first approach would involve a route-to-route extrapolation of the inhalation data. The occupational exposure study of Bättig et al. (1958) was used to derive a provisional RfC for both 1,2,4-TMB and 1,3,5-TMB. This study, however, has several limitations in that the workers were exposed to a mixture consisting of 50% 1,2,4-TMB and 30% 1,3,5-TMB; the composition of the remaining 20% of the mixture was unknown. The length of exposure and the concentrations were not well defined. With these limitations, added to the additional uncertainty introduced by route-to-route extrapolation, derivation of an RfD based on the inhalation database is not recommended. The second approach involves deriving an RfD for 1,2,4-TMB by analogy to 1,3,5-TMB.

There are little data to support or refute deriving an RfD for 1,2,4-TMB by analogy to 1,3,5-TMB. Mikulski and Wiglusz (1975) found that the excretion of TMB metabolites was similar (in terms of types of metabolites and kinetic constants) in rats administered a single oral dose of 1200 mg/kg 1,2,4- or 1,3,5-TMB. The primary urinary metabolites of 1,2,4-TMB and 1,3,5-TMB are 3,4-dimethylhippuric acid and 3,5-dimethylhippuric acid (Mikulski and Wiglusz, 1975), respectively. Similarly several recent inhalation toxicokinetic studies in humans have shown that the elimination kinetics and metabolism of 1,2,4-TMB and 1,3,5-TMB in humans are also similar (Jarnberg, et al., 1996, 1997; Kostrewski, et al., 1997).

If the assumption is made that 1,2,4- and 1,3,5-TMB are similarly metabolized (as supported by urinary excretion data) and have similar toxicological endpoints (potency and target organs), then an RfD for 1,2,4-TMB can be derived by analogy to 1,3,5-TMB. The similarity in the structure of 1,2,4- and 1,3,5-TMBs suggests that the two isomers can be used as surrogates for the other. In a chronic oral exposure study, Maltoni et al. (1997) reported a slight to moderate decline in survival of rats after 104 weeks of exposure to an adjusted dose of 457 mg/kg-day. This exposure level greatly exceeds the subchronic NOAEL level of 143 mg/kg-day used to derive an RfD for 1,3,5-TMB. Thus, using the lower NOAEL level from the IITRI (1995) 1,3,5-TMB study to derive a provisional RfD for 1,2,4-TMB should be adequately conservative and protective.

The provisional RfD is derived using the NOAEL of 200 mg/kg (143 mg/kg-day) identified in the IITRI (1995) study in which rats received gavage dosages of 1,3,5-TMB 5 days/week for 90 days. This NOAEL is divided by an uncertainty factor of 3000 (10 to account for using a less-than-lifetime study, 10 for interspecies extrapolation, 10 for human variability, and 3 for database deficiencies). The resultant provisional RfD for 1,2,4-TMB is 5E-2 mg/kg-day. The database limitations include lack of oral toxicity studies for 1,2,4-TMB and developmental and reproductive toxicity studies. Confidence in the principal study (IITRI, 1995) is high; it is a well-designed study examining a number of relevant endpoints. Confidence in the database is low; no oral toxicity study examining a number of relevant endpoints was located for 1,2,4-TMB, the database for 1,3,5-TMB lacks an oral toxicity study in a second species, and no information is available on the potential of ingested 1,3,5-TMB to induce developmental, reproductive, or neurological effects (a sensitive endpoint in inhalation studies). Reflecting the low confidence in the database, confidence in the RfD is low.

RISK CHARACTERIZATION

The TMB isomers, 1,2,4-TMB and 1,3,5-TMB, have not been studied for their carcinogenic potential in humans. Only one long term oral exposure animal study has been performed to assess the carcinogenic potential of 1,2,4-TMB in Sprague-Dawley rats (Maltoni et al., 1977). There was no significant increase in the incidence of animals bearing either malignant or benign + malignant tumors after exposure to an adjusted dose of 457 mg/kg-day for 104 weeks. The incidence of neuroesthesioepitheliomas in the nasal cavity of exposed animals (M + F) was 3%, as opposed to 0% in the control animals. The increase in incidence of neuroesthesioepitheliomas, however, was not statistically significant (p<0.05). This study was conducted only with one species, rats, and with only one dose level, thus it is possible that 1,2,4-TMB could be carcinogenic in a different species or at a different dose level. The study was inadequate to establish clear evidence that 1,2,4-TMB is noncarcinogenic. Both 1,2,4-TMB and 1,3,5-TMB were found to be negative for mutagenicity in the Ames test, negative in the mouse micronucleus test, and positive in the mouse SCE test (Janik-Spiechowicz et al., 1998). These data were considered to provide inadequate evidence for genotoxic activity.

The human carcinogenicity potential of 1,2,4-TMB and 1,3,5-TMB cannot be determined on the basis of the available information. Both human and animal data are judged inadequate for an evaluation. However, the Maltoni study provides suggestive evidence of potential carcinogenicity, but there is not definitive evidence. Additional studies are needed for a full evaluation of the potential carcinogenicity of 1,2,4-TMB and 1,3,5-TMB. Hence following the 1996, proposed guidelines for carcinogen risk assessment, a dose response assessment is not appropriate (U.S. EPA, 1996).

ACGIH (American Conference of Governmental Industrial Hygienists). 1998. 1998 Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices. ACGIH, Cincinnati, OH. p. 69.

ATSDR (Agency for Toxic Substances and Disease Registry). 1995. Toxicological Profile for Xylenes. U.S. Department of Health & Human Services, Public Health Service, Atlanta, GA.

Bättig, K., E. Grandjean, L. Rossi and J. Rickenbacher. 1958. Toxicological investigation of trimethylbenzene. Arch. Gewerbephatol. Gewerbehygiene. 16: 555-566. [German].

Borriston Laboratories, Inc. 1984. Four-week oral nephrotoxicity screening study in male F344 rats. EPA Document No. FYI-AX-0784-0280.

HSDB (Hazardous Substances Data Bank). 1998. Online. National Library of Medicine. Washington, DC.

IITRI (IIT Research Institute). 1995. 90-Day oral gavage toxicity study of 1,3,5-trimethylbenzene in rats with a recovery group. Conducted for Koch Industries Inc., Wichita KS.

Janik-Spiechowicz, E., K. Wyszska and E. Dziubaltowska. 1998. Genotoxicity evaluation of trimethylbenzenes. Mut. Res. 412: 299-305.

Järnberg, J., G. Johanson and A. Löf. 1996. Toxicokinetics of inhaled trimethylbenzenes in man. Toxicol. Appl. Pharmacol. 140: 281-288.

Järnberg, J., B. Ståhlbom and G. Johanson. 1997. Urinary excretion of dimethylhippuric acids in humans after exposure to trimethylbenzenes. Int. Arch. Occup. Environ. Health. 69: 491-497.

Kostrewski, P., A. Wiaderna-Brycht and B. Czerski. 1997. Biological monitoring of experimental human exposure to trimethylbenzene. Sci. Tot. Environ. 199: 73-81.

Maltoni, C., A. Ciliberti, C. Pinto, M. Soffritti, F. Belpoggi and L. Menarini. 1997. Results of long-term experimental carcinogenicity studies of the effects of gasoline, correlated fuels, and major gasoline aromatics on rats. Ann. New York Acad. Sci. 837: 15-52.

Mikulski, P.I. and R. Wiglusz. 1975. The comparative metabolism of mesitylene, pseudocumene, and hemimellitene in rats. Toxicol. Appl. Pharmacol. 31: 21-31.

NIOSH (National Institute for Occupational Safety and Health). 1994. Pocket Guide to Chemical Hazards. U.S. Department of Health and Human Services, Washington, DC. p. 64.

OSHA (Occupational Safety and Health Administration). 1997. 29 CFR 1910.1000. Office of Federal Register, National Archives and Records Administration. Washington, DC.

Pyykkö, K. 1980. Effects of methylbenzenes on microsomal enzymes in rat liver, kidney and lung. Biochem. Biophys. Acta. 633: 1-9.

U.S. EPA. 1987a. Health Effects Assessment for Trimethylbenzenes. Prepared by the Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, DC.

U.S. EPA. 1987b. Drinking Water Health Advisory for 1,2,4-Trimethylbenzene. Prepared by the Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment, Cincinnati, OH for the Office of Drinking Water, Washington, DC.

U.S. EPA. 1994. Office of Health and Environmental Assessment Chemical Assessments and Related Activities. Office of Health and Environmental Assessment, Washington, DC. December, 1994. OHEA-I-127.

U.S. EPA. 1996. Proposed Guidelines for Carcinogen Risk Assessment. Federal Register 61:17960-18011.

U.S. EPA. 1997. Health Effects Assessment Summary Tables. FY-1997 Update. Office of Research and Development, Office of Solid Waste and Emergency Response, Washington, DC. OSWER Dir. 9200-6-303 (97-1).

U.S. EPA. 1998. Integrated Risk Information System. Online. Office of Research and Development, National Center for Environmental Assessment, Cincinnati, OH.