Determination of urinary 2,5-hexanedione concentration by an improved analytical method as

Transcription

1 568 British Journal of Industrial Mediine 1991;48: Determination of urinary 2,5-hexanedione onentration by an improved analytial method as an index of exposure to n-hexane Isao Saito, Eiji Shibata, Jian Huang, Naomi Hisanaga, Yuihiro Ono, Yasuhiro Takeuhi Abstrat 2,5-Hexanedione is a main metabolite of n-hexane and is onsidered as the ause of n-hexane polyneuropathy. Therefore, it is useful to measure 2,5-hexanedione for biologial monitoring of exposure to n-hexane. The analytial methods existing for n-hexane metabolites, however, were ontroversial and not established enough. Hene, a simple and preise method for determination of urinary 2,5-hexanedione has been developed. Five ml of urine was aidified to ph -5 with onentrated hydrohlori aid and heated for 3 minutes at 9-1. After ooling in water, sodium hloride and dihloromethane ontaining internal standard were added. The sample was shaken and entrifuged. 2,5-Hexanedione onentration in an aliquot of dihloromethane extrat was quantified by gas hromatography using a widebore olumn (DB-171). Urinary onentration of 2,5-hexanedione showed a good orrelation with exposure to n-hexane (n = 5, r = -973, p < -1). This method is simple and preise for analysis of urinary 2,5-hexanedione as an index of exposure to n- hexane. n-hexane is a widely used solvent in industry and well known to ause polyneuropathy.'-' It is metabolised to 2,5-hexanedione, 2,5-dimethylfuran, 2-hexanol, and o-valerolatone. 2,5-Hexanedione, whih is one of the main metabolites of n-hexane, is onsidered as the ultimate neurotoxi metabolite.' Many investigations into the relation between urinary metabolites and exposure to n-hexane have been Aihi Prefetural Institute of Publi Health, Tsujimahi, Kita-ku, Nagoya 462 Japan Isao Saito Department of Hygiene, Nagoya University Shool of Mediine, Showa-ku, Nagoya 466 Japan E Shibata, J Huang, N Hisanaga, Y Ono, Y Takeuhi arried out and good orrelations have been found between exposure to n-hexane and urinary 2,5- hexanedione onentration.7"' These results suggest that measurement of urinary 2,5-hexanedione onentration ould be used for biologial monitoring of exposure to n-hexane. Fedtke and Bolt showed that 4,5-dihydroxy-2-hexanone was a main urinary metabolite of n-hexane but this metabolite was onverted to 2,5-hexanedione by aid hydrolysis.'2 Fifty ppm is proposed as the eight hour time weighted average threshold limit value (TLV-TWA) of n-hexane by The Amerian onferene of Governmental Industrial Hygienists (AGIH)" and 4 ppm by The Japan Assoiation of Industrial Health.'4 As a biologial exposure index (BEI), 5- mg/l of urinary 2,5-hexanedione is proposed by AGIH'3 and 9 mg/l by Deutshe Forshungsgemeinshaft (DFG).'s In Japan we have a legal obligation sine 1989 to measure 2,5-hexanedione in the urine of workers exposed to n-hexane. The exposed workers are lassified into three groups by the onentration of urinary 2,5-hexanedione. The first group has a onentration below 2 mg/l of 2,5- hexanedione, the seond group from 2 to 5 mg/l, and the third group more than 5 mg/l. The analytial method devised by Perbellini et al uses enzymati and aid hydrolysis and double extration with dihloromethane and evaporation of the solvent with nitrogen gas.'6 Urinary metabolites of n-hexane have been mostly measured by the method of Perbellini et al'6 or slight modifiations of this. It is time onsuming, however, and some loss of 2,5-hexanedione ours during onentration. The method of Fedtke and Bolt'7 is simpler but it needs neutralisation to ph 7 after aid hydrolysis and is less sensitive than the method of Perbellini et al 6 beause the urine sampling volume is limited to only 1 ml. Reently, improved methods for the determination of 2,5- hexanedione have been reported,'819 but few fundamental investigations have been made into the analytial onditions neessary for aurate and sensitive measurement of 2,5-hexanedione. The present study aimed to develop a simple and preise method for analysis of urinary 2,5-hexanedione.

2 Determination of urinary 2,5-hexanedione onentration as an index of exposure to n-hexane Materials and methods HEMIALS All reagents and hemials were speial or first grade, obtained from Tokyo Kasei Industry (Tokyo, Japan), Katayama hemial (Osaka, Japan), and Wako Pure hemial Industry (Osaka, Japan). INSTRUMENTATION Urine analyses were determined on a Shimadzu G-R1A or G-15A gas hromatograph equipped with flame ionisation detetor. olumns were fused silia widebore (megabore) -53 mm internal diameter x 15 m DB-171 (14% yanopropyl phenyl polysiloxane), DB-1 (dimethyl polysiloxane), DB-5 (5% diphenyl and 95% dimethyl polysiloxane), DB-17 (5% diphenyl and 5% dimethyl polysiloxane), DB-21 (5% trifluoropropyl and 5% methyl polysiloxane), and DB-WAX (polyethyleneglyol 2 M) all from J andw Sientifi (A, USA), and apillary olumns -32 mm internal diameter x 3 m BP-1 (equivalent to DB-1) and BP-2 (equivalent to DB-WAX) from Shimadzu (Kyoto, Japan). Helium was used as the arrier gas at a flow rate of 15 ml/min for a widebore olumn and 2 ml/min for a apillary olumn. The temperatures of injetor and detetor were 12 and 2 respetively. The olumn temperature was programmed to be 5 for the initial three minutes, then inreased from 5 to 1 at 5 /min; subsequently from 1 to 2 at 2 /min to purge the omponent extrated from urine. The gas hromatography-mass spetrometry analyses were performed with wide bore olumn DB-171 and JEOL JMS-AX33 with an eletron energy of 7 ev. URINE PROESSING Urine samples were olleted from 5 workers (27 women and 23 men) painting and spraying adhesives for ar interiors in four fatories. The samples were olleted at the end of the workshift and kept at -4 until analysis. Urine was also olleted from Wister strain male rats (body weight 26-3 g) exposed to 2 ppm of n-hexane ontinuously for 12 h/day. AIR SAMPLING The exposure onentration of n-hexane and other solvents were measured by a Pro Tek personal monitoring badge (Du Pont, USA). The badge was attahed to the worker's ollar during the workshift. After use, the badge was sealed tightly in a small bag and kept at 4 until analysis. The arbon felt in the badge was steeped in S2 for one hour and the solvent was injeted into the gas hromatograph (olumn BP-1 or BP-2). Eight hour time weighted average onentrations (TWA) were alulated. URINE ANALYSIS A 5 ml urine sample was plaed in a glass tube (1 ml volume) with a srew ap. onentrated hydrohlori aid (11 3 mole; 2 ml) was added. The mixture was heated for 3 minutes in a water bath (9-1 ), then ooled with water. Sodium hloride (1 5 g) and 1 ml of dihloromethane ontaining 2 nl/ml eah of 3-methyl-ylohexanone and ylohexanone as internal standards were added. The sample was shaken vigorously for three minutes and entrifuged at 3 rpm for five minutes. The dihloromethane layer was separated and dried with anhydrous Na2SO4 and 2 p1 of the solution was injeted into the gas hromatograph. The onentration of 2,5-hexanedione was orreted to a speifi gravity of For a alibration urve, 2 5, 5, 1, 15, and 25 pg of standard solution of 2,5-hexanedione were added to 5 ml water and 1-5 g of sodium hloride was added. These solutions were extrated with dihloromethane and analysed as above (solutions are equivalent to, 5, 1, 2, 3, and 5 mg/l). Results GAS HROMATOGRAPHI ONDITIONS Extrats of urine from non-exposed persons were measured using various gas hromatographi olumns. The table shows the results. The gas hromatograms of 2,5-hexanedione obtained from DB-171, DB-1, and DB-5 olumns showed less interferene than those obtained from polar olumns. Subsequently, therefore, urinary 2,5-hexanedione was determined mainly using the wide bore olumn omparison with various gas hromatography olumns of onentration of 2,5-hexanedione in the urine of workers not exposed to n-hexane olumn type DB-1 71 DB-1 DB-5 DB-17 DB-21 DB-WAX BP-1 BP-2 2,5-Hexanedione (mg/l)* Range (mg/i) oeffiient of variation (%) The onentration of 2,5-hexanedione was alulated from the peak with the same retention time as standard 2,5-hexanedione. *These values are mean (SD) of 1 urine samples. 569

3 57 E6.\ 3 x I 2- [Lontrol.2~V ph Figure 1 Effet of hange in ph on 2,5-hexanedione onentration in urine from workers and rats exposed to n-hexane. DB-171. ylohexanone is used as an internal standard in gas hromatography, but in some ases an overlapping peak was observed. As well as ylohexanone, we seleted 3-methyl ylohexanone as an internal standard. This has a retention time near to that of 2,5-hexanedione and few overlapping interfering ompounds. For the determination of 2,5- hexanedione the oeffiient of variation by the internal standard method was three times more preise than that by the diret method. The detetion limit of 2,5-hexanedione in urine was mg/l by extrating 5 ml of urine in the present method. AID HYDROLYSIS ONDITIONS We studied optimal ph onditions for aid hydrolysis with urine of workers exposed and not exposed to n-hexane, and rat urine. Figure 1 shows that the amount of 2,5-hexanedione depended on the adjusted ph in urine before heating the sample. 2,5-Hexanedione onentrations in urine of workers exposed to n-hexane were onsistently low from ph 5 to ph 3. 2,5-Hexanedione onentration rapidly inreased, however, from ph 3 to ph 1- and reahed a plateau at ph 1. The urve for 2,5-hexanedione with hanging ph in the urine of workers is similar to that for rat urine. 2,5- Hexanedione in the urine of workers not exposed to n-hexane was not inreased even below ph 2-. Saito, Shibata, Huang, Hisanaga, Ono, Takeuhi Therefore, ph for aid hydrolysis was adjusted to around ph 5 by adding 2 ml onentrated hydrohlori aid. For 4 workers, the mean ph of the urine, to whih -2 ml of onentrated hydrohlori aid was added was 49 (SD 7), range 3-7. EFFET OF ph ON THE STABILITY OF URINARY METABOLITES The effet of ph on the stability of 2,5-hexanedione, 2,5-dimethylfuran, and 2-hexanol, (urinary metabolites of n-hexane) was studied. Even below ph 2- the amount of 2,5-hexanedione and 2-hexanol were not hanged. Figure 2 shows, however, that 2,5-dimethylfuran onentration was rapidly dereased below ph 3- and disappeared below ph 2-. On the other hand, the peak of 2,5-hexanedione appeared below ph 3. Therefore, after aid treatment to give a ph of less than 1 the urinary metabolite 2,5-dimethylfuran was onsidered to be totally onverted to 2,5-hexanedione and determined as 2,5-hexanedlone in the present method. EFFET OF TIME AND TEMPERATURE ON AID HYDROLYSIS We studied the effet of temperature (6-1 ) and heating time (-6 minutes) on aid hydrolysis. The amount of 2,5-hexanedione in the urine inreased with inreases in temperature and heating time and reahed a maximum at over 8 and 15 minutes. The onditions for aid hydrolysis, therefore, were hosen to be 9-1 and 3 minutes. EXTRATION OF 2,5-HEXANEDIONE FROM HYDROLYSED URINE We investigated an effetive and simple extration method. The extration effiienies of five solvents (dihloromethane, ethyl aetate, diethyl ether, isopropyl ether, and n-pentane) were ompared. All o > _ S.2 a 2 ph Figure 2 Effet ofph on the onversion of 2,5-dimethylfuran to 2,5-hexanedione. 2,5 - Dimethylfuran 2,5- Hexanedione 4

4 Determination of urinary 2,5-hexanedione onentration as an index of exposure to n-hexane 1 15 Retention time (min) Figure 3 Gas hromatograms of the urine of workers exposed and not exposed to n-hexane, using the proposed method: (a) the urine of worker exposed to n-hexane; (b) the urine of worker not exposed to n-hexane; () standard solution (peak 1, ylohexanone, peak 2, 3-methyl ylohexanone, peak 3, 2,5-hexanedione). the solvents had good extration effiienies for 2-hexanol and 2,5-dimethylfuran. Dihloromethane, however, had greater effiieny for 2,5-hexanedione than the other solvents and we hose this as the extrating solvent. Using 1 ml of dihloromethane to extrat from 5 ml of aqueous solutions the extration effiieny was about 95% for 3-methyl ylohexanone and ylohexanone used as internal standard but about 7% for 2,5-hexanedione. Therefore, the effet of sodium hloride on extration effiieny of 2,5-hexanedione was studied. The extration effiieny was improved to about 9% by adding 1 5 g of sodium hloride. REOVERY OF 2,5-HEXANEDIONE ADDED TO HUMAN URINE Reovery of 5 mg 2,5-hexanedione added to urine was determined five times. Mean reovery was 86-2 (SD 1-3)%. The reovery of 2,5-hexanedione was improved to 97-6 (SD 1-7; V 1-8)%, however, by omparison with a alibration urve made by extration of an aqueous solution of 2,5-hexanedione in the same way as the urine sample. IDENTIFIATION OF URINARY 2,5-HEXANEDIONE Figure 3 shows a gas hromatogram of the extrated urine of non-exposed workers and workers exposed to n-hexane, with the proposed method. The peak of 2,5-hexanedione in urine samples from workers exposed to n-hexane was seen at the same retention time as standard 2,5-hexanedione. The mass spetrum of this urinary peak was idential with that of standard 2,5-hexanedione (fig 4). STABILITY OF 2,5-HEXANEDIONE DURING STORAGE We studied the stability of 2,5-hexanedione using urine from rats exposed to n-hexane, diluted with human urine. The urine mixture was allowed to stand at 25, 4, or - 2 for 3 days. Urinary 2,5-hexanedione onentration at -2 and 4 was unhanged by 3 days. At 25, however, urinary 2,5-hexanedione onentration was less after one day and was one fifth of the initial onentration after 15 days. 'ao a 1-8^ 6I :57 1 H3 H2 H2.H3 III II I Relative mass Figure 4 Mass spetrum of 2,5-hexanedione: (A) standard solution; (B) urine of worker exposed to n-hexane. 99,99 ) 571 M+ 114.I 1 12o

5 572 33, y=5x+-8 r=-973 (p<-1) X 2 W x I ^ TWA onentration of n - hexane (ppm) Figure 5 Relation between exposure onentration and 2,5-hexanedione onentration in urine of workers. RELATION BETWEEN THE EXPOSURE ONENTRATION OF N-HEXANE AND URINARY 2,5-HEXANEDIONE Mean TWA onentrations of solvents at the four fatories were: fatory 1 (n = 2), n-hexane 2-6 (SD 3-6) ppm, toluene 24-6 (SD 41-8) ppm, and MEK 29-2 (SD 26 6) ppm; fatory 2 (n = 4), n-hexane 28-2 (SD 6 8) ppm and toluene 22-2 (SD 4 8) ppm; fatory 3 (n = 4), n-hexane -1 (SD -1) ppm and toluene 2 5 (SD 1-2) ppm; fatory 4 (n = 22), n-hexane 3-6 (SD 2 7) ppm and toluene 6-7 (SD 4 ) ppm. 2,5- Hexanedione onentrations in the urine of 5 workers were determined by the present method. Figure 5 shows the good orrelation between exposure onentration of n-hexane and urinary 2,5-hexanedione. Disussion The environmental onentration of organi solvent is useful for estimating exposure, but it does not always aurately reflet the intake of the individual worker. Therefore, biologial monitoring using urinary metabolites is an important method for the atual assessment of exposure to organi solvents. Many investigations show that urinary 2,5-hexanedione is suitable as an indiator of exposure to n-hexane.7"l Perbellini et al'622' have investigated the urinary metabolites of n-hexane. In 198 they reported that 2,5-dimethylfuran, 2-hexanol, 2,5-hexanedione, and y-valerolatone were deteted in the urine of workers exposed to n-hexane in shoe fatories. The method used enzymati and aid hydrolysis (ph 2 ), double extration with dihloromethane, and onentration by evaporation with a stream of nitrogen gas. This method is, however, time onsuming for the quantitative analysis of urinary n-hexane metabolites. Fedtke and Bolt developed a simple method using Saito, Shibata, Huang, Hisanaga, Ono, Takeuhi aid hydrolysis (ph -1) and artridges for extration of n-hexane metabolites in 1 ml urine.22 Their investigations of urinary n-hexane metabolites showed that the onjugated preursor that onverted to 2,5- hexanedione after aid hydrolysis was 4,5- dihydroxy-2-hexanone and that the enzymati hydrolysis step was not neessary. " During urine analysis, this metabolite is onverted to 2,5- dimethylfuran by weak aid hydrolysis and to 2,5- hexanedione by strong aid hydrolysis. Urinary 2,5- hexanedione is a useful indiator for biologial monitoring, however, beause it orrelates well with exposure to n-hexane. In our present study 2,5-dimethylfuran was found to onvert to 2,5- hexanedione below ph 1-. Therefore, 4,5- dihydroxy-2-hexanone and 2,5-dimethylfuran as metabolites of n-hexane in the urine of workers exposed to n-hexane ould be measured as 2,5- hexanedione by hydrolysing below ph 1-. Sampling volumes of urine for analysis were 1 ml (Perbellini et al'6) and 1 ml (Fedtke and Bolt22). The detetion limit of2,5-hexanedione was generally 5-1- ng by gas hromatography. Therefore, we seleted 5 ml as a sampling volume of urine, by whih the detetion limit for measurement of 2,5-hexanedione onentration in urine was -5-O 1 mg/l. For the determination of 2,5-hexanedione, Perbellini et al"6 inubated the urine sample at ph 4-8 with /3- gluuronidase for 24 hours and then heated it in boiling water at ph 2- for 3 minutes. Fedtke and Bolt hydrolysed the urine sample in boiling water at ph -1 for 3 minutes. We seleted ph 5 beause aid hydrolysis at ph 2- does not ause maximum release of 2,5-hexanedione and 2,5-hexanedione does not break down between ph -1 and ph.5. Furthermore, the interfering peaks in gas hromatograms were less at ph 5 than at ph -1. We also studied the effet of temperature and time on the aid hydrolysis. The amount of 2,5-hexanedione after aid hydrolysis reahed a maximum at over 8 and by 15 minutes. Therefore, we hose the ondition at 9-1 and 3 minutes for aid hydrolysis. For the extration of 2,5-hexanedione from hydrolysed urine, dihloromethane has been traditionally used as an extrating solvent. Reently, the fat that dihloromethane has a weak arinogeniity was reported.24 Therefore, we studied other extrating solvents that had good extratability and were safer than dihloromethane. Dihloromethane was found to be the best extrating solvent, however, for urinary n-hexane metabolites so we used small amounts of dihloromethane as the extrating solvent and did not onentrate the extrat. About 7% of 2,5-hexanedione in 5 ml of aqueous solution was extrated with 1 ml of dihloromethane. The addition of sodium hloride improves, the reovery of 2,5-hexanedione from the solution after

6 Determination of urinary 2,5-hexanedione onentration as an index of exposure to n-hexane 573 aid hydrolysis by some 2%. When 2,5-hexanedione solution was added to two urine samples (speifi gravity 1-31 and 1-8) after aid hydrolysis, the reoveries of 2,5-hexanedione extrated with dihloromethane were 77% and 72% respetively. The differene in the reovery was not apparent after adding sodium hloride to the urine. Therefore, the addition of sodium hloride might also redue the variation in reovery due to differenes in speifi gravity of urine. We have heked various types of gas hromatography olumns to separate the peak of 2,5- hexanedione from interfering peaks. The interfering peaks did not appear when using non-polar or weak polar olumns and we seleted olumns DB- 171 and DB-1 for the analysis. Fedtke and Bolt22 showed that the amount of 2,5- hexanedione deteted in urine of workers not exposed to n-hexane was 45 (SD 2) mg/l (n = 12, range mg/l). Our results were -12 (SD 5) mg/l (n = 1, range < 1-2 mg/l), whih were lose to the detetion limits on analysing by DB-171 olumn. The differene between these results might be due to the differene of ph on aid hydrolysis. The reason is not learly understood, but at least the onentrations of urinary 2,5-hexanedione in non-exposed persons measured by the proposed method were near the detetion limit. The orrelation between level of n-hexane exposure and 2,5-hexanedione onentration in the urine of 5 workers at the end of the workshift is good at low onentrations of n-hexane. The onentration of urinary 2,5-hexanedione orresponding to an n-hexane exposure onentration of 5 ppm was 2-7 mg/l in the present study. Other authors reported 5-4 mg/i,7 2-7 mg/i,8 3-2 mg/i,9 4-2 mg/l,' and 1-6 mg/l. " We think that these differenes may be due to the differene in method of analysis. In these reports 2,5-hexanedione was determined by the method of Perbellini et al'6 or modifiations of this. Gas hromatographi olumns used for determining 2,5-hexanedione were usually polar (polyethylene glyol 2 M (arbowax 2 M, PEG2 M, DB- WAX, BP-2 et)), although Iwata et al82526 used a non-polar olumn OV-11. The mixed exposure with other solvents suh as toluene and MEK might have some effets on the exretion of urinary 2,5- hexanedione Mean TWA onentrations of toluene and MEK in fatory 1 were 24-6 and 29-1 ppm, respetively but the onentration of n- hexane (2-6 ppm) was too low to allow the assessment of the effet of other solvents on the exretion of urinary 2,5-hexanedione. Mean TWA onentrations of n-hexane and toluene in fatory 2 were 28-1 and 22-2 ppm, respetively, however, and a definite effet of other solvents on onentration of 2,5- hexanedione was not suggested in the present study. These problems will be resolved by ongoing investigations with mixed solvents using the same analytial method. In summary, in this study, we have developed a simpler and more preise analytial method of urinary 2,5-hexanedione than previous methods. Some similarities exist with two other reent reports of improved methods.'8 19 Our present method showed a high orrelation between urinary 2,5-hexanedione and exposure to n-hexane. The method will be useful for biologial monitoring of exposure to n-hexane. 1 Wada Y, Okamoto S, Takagi S. Intoxiation polyneuropathy following exposure to n-hexane. lin Neurol Neurosurg 1965;5: Yamada S. Intoxiation polyneuropathies in the workers exposed to n-hexane. Japanese Journal of Industrial Health 1967;9: Herskowitz A, Ishii N, Shaumburg HH. n-hexane neuropathy. A syndrome ourring as a result of industrial exposure. N Engl J Med 1971;285: Krasavage WJ, O'Donoghue JL, DiVinenzo GD, Terhaar. The relative neurotoxiity of methyl-n-butyl ketone, n- hexane and their metabolites. Toxiol Appl Pharmaol 198;52: ouri D, Milks M. Toxiity and metabolism of the neurotoxi hexaarbons n-hexane, 2-hexanone, and 2,5-hexanedione. Annual Review of Pharmaology and Toxiology 1982;22: Governa M, aliste R, oppa G, Tagliavento G, olombe A, Troni W. Urinary exretion of 2,5-hexanedione and peripheral polyneuropathies in workers exposed to hexane. J Toxiol Environ Health 1987;2: Perbellini L, Brugnone F, Faggionato G. Urinarv exretion of the metabolites of n-hexane and its isomers during oupational exposure. Br J Ind Med 1981;38: Iwata M, Takeuhi Y, Hisanaga N, Ono Y. A study on biologial monitoring of n-hexane exposure. Int Arh Oiup Environ Health 1983;51: Mutti A, Falzoi M, Luertini S, et al. n-hexane metabolism in oupationally exposed workers. Br J Imrd Med 1984;41: De Rosa E, Bartolui GB, Perbellini L, Brugnone F, Rausa G. Environmental and biologial monitoring of exposure to toluene, styrene, and n-hexane. Applied Industrial Hygiene 1988;3: Ahonen I, Shimberg RW. 2,5-Hexanedione exretion after oupational exposure to n-hexane. Br J Id Ated 1988; 45: Fedtke N, Bolt HM. 4,5-Dihydroxy-2-hexanone. A new metabolite of n-hexane and of 2,5-hexanedione in rat urine. Biomed Environ Mass Spetronit 1987;14: Amerian onferene of Governmental Industrial Hygienists. Threshold limit values and biologial exposure indiesfor ininnati, Ohio: AGIH, The Japan Assoiation of Industrial Health: Reommendations of TLVs. Japanese Journal of Industrial Health 1989;31: Deutshe Forshungsgemeinshaft. MaximuOn onentrations at the workplae and biologial toleranie values for working materials New York: VH Publishers, Perbellini L, Brugnone F, Silvestri R, Gaffuri E. Measurement of the urinary metabolites of n-hexane, ylohexane and their isomers by gas hromatography. Int Arh Oup Environ Health 1981;48: Fedtke N, Bolt HM. Methodologial investigations on the determination of n-hexane metabolites in urine. Int Arh Oup Environ Health 1986;57: Perbellini L, Amoros DM, Llorens A, Giuliani, Brugnone F. An improved method of analysing 2,5-hexanedione in urine. Br J Ind Med 199;47: Kawai T, Mizunuma K, Yasugi T, Uhida Y, Ikeda M. The method of hoie for the determination of 2,5-hexanedione as an indiator of oupational exposure to n-hexane. Int Arh Oup Environ Health 199;62: Perbellini L, Brugnone F, Pavan I. Identifiation of the metabolites of n-hexane, ylohexane, and their isomers in men's urine. Toxiol Appl Pharmaol 198;53: Perbellini L, Amantini M, Brugnone F, Frontali N. Urinary

7 574 exretson ot n-hexane metabolites. Arh Toxiol 1982;5: Fedtke N, Bolt HM. Detetion of 2,5-hexanedione in the urine of persons not exposed to n-hexane. Int Arh Oup Environ Health 1986;57: Fedtke N, Bolt HM. The relevane of 4,5-dihydroxy-2- hexanone in the exretion kinetis of n-hexane metabolites in rat and man. Arh Toxiol 1987;61: International Ageny for Researh on aner. Monograph on the evaluation of the arinogeni risk of hemials to humans. Vol 41. Lyon: IAR, 1986: Iwata M, Takeuhi Y, Hisanaga N, Ono Y. hanges of n-hexane metabolites in urine of rats exposed to various onentrations Saito, Shibata, Huang, Hisanaga, Ono, Takeuhi of n-hexane and to its mixture with toluene or MEK. Int Arh Oup Environ Health 1983;53: Iwata M, Takeuhi Y, Hisanaga N, Ono Y. hanges ofn-hexane neurotoxiity and its urinary metabolites by long-term oexposure with MEK or toluene. Int Arh Oup Environ Health 1984;54: Shibata E, Huang J, Ono Y, et al. hanges in urinary n-hexane metabolites by o-exposure to various onentrations of methyl ethyl ketone and fixed n-hexane levels. Arh Toxiol 199;64: Aepted 17 Deember 199 Br J Ind Med: first published as /oem on 1 August Downloaded from on 3 September 218 by guest. Proteted by