Document GbKK2RmnwrMjaVJ4wMxbpJL7

Occupational and Environmental Medicine Papers 577 581 598 602 606 612 618 625 632 640 Correspondence 646 646 647 647 647 Lymphohaematopoieticmalignancy around all industrial complexes that include major oil refineries in Great Britain P Wilkinson,B Thakrar,P Walls,M Landon, S Falcom, C Gmndy, P Elliott Mortality among aircraft manufacturing workers ]ohn D Boice Jr, Donald E Marano, Jon P Fryzek, ChristopherJ Sadlet, JosephK McLuugblin Biological monitoringto assess exposure from use of isocyanates in motor vehicle repair N R William, K ]ones, J Cocker Asbestosis and small cell lung cancer in a clutch refabricator JeffreyL b i n , Michael F O'Sullivan, CarolynJ Corn, M G h William, Ronald F D d o n Chemical differencesbetween long and short amosite asbestos: differencesin oxidation state and coordination sites of iron, detected by infra red spectroscopy Andrew Graham,John Higin6otbam, Doug Alkan, Kenneth Donaldson, Paul H Be&& Non-neoplastic mortality of European workers who produce man made vitreous fibres D Sali, P Boffetta, A Andersen, J W Cberrie, J Cbang Claude, J Hatuen, J H Oken, A C Pmatori, N Pkato, L T w o , P Westerholm,P Wintq R Saraui Study of cancer incidence among 6363male workers in four Nomyfercomangancse and silicomvlganeseproducing plants nund H066esland, Helge Kjuus,Dag S Thelk Study of cancer incidence among 8530 male workers in eight Norwegian plants producing ferrosilicoa and silicon metal Anumi Hob6csland, Helge Kjuus,Dag S Thelk Decrements in cognitiveperformance in metal inert gas welders exposed to aluminium &a Akika, Brian T Stolkry, Vesa Riihimiiki Shift work, social class, and ischaemic heart disease in middle aged and elderly men; a 22 year follow up in the Copenhagen male study Henrik Beggdd, Poul Suudicuni, Hans Ok Hein, Finn Gyntelberg Basal cell carcinoma after frequent episodes of cutaneous erythema and peeling induced by welding A Michael Donoghue, Murray J Sinckair Work related respiratory symptoms in radiographers Julia Smedky, Paul Cullinan,Anthony Frew, Anthony Newman-Tqloc David Coggon Air pollution research in the United Kingdom R L Maynard Notice Book reviews . . -. Lymphohaematopoietic malignancy around all industrial complexes that include major oil refineries in Great Britain P Wilkinson, B Thakrar, P Walls, M Landon, S Falconer, C Grundy, P Elliott Abstract crease for lymphohaematopoietic malignancy.` Objectives--To examine the incidence Another study in the United States suggested of lymphohaematopoietic malignancy high rates of lung cancer near petrochemical around industrial complexes that include plant^.^ A study of the geographical disuibu- major oil refineries in Great Britain after tion of childhood leukaemia and other child- recent public and scientific concern of hood cancers in Great Britain reported a possible carcinogenic hazards of emis- tendency for clustering near emission sources sions from the petrochemical industry. of (among other things) volatile petroleum Methods-Small area study of the inci- products,6 although the methodology of this dence of lymphohaematopoietic malig- study with residential post codes as a proxy for nancies, 1974-91, within 7.5 k m ofall 11oil populations had previously been questioned.' refineries (grouped into seven sites) in A more recent study of the same data set simi- Great Britain that were operational by the larly found an association of childhood cancers early 1970s and processed more than two with emission sources based on comparisons million tonnes of crude oil in 1993. between address at birth and at death.' By con- t Results-Combined analysis of data from trast, studies of cancer incidence and mortality all seven sites showed no significant among residents living close to a petrochemical (pcO.05) increase in risk of these malig- plant in south Wales were largely rea~suringl.o~ nancies within 2 km or 7.5 km.Hodgkin's lymphoma, but no other malignancy, showed evidence (p=0.02) of a decline in risk with distance &om refineries, but We report here a study of lymphohaematopoietic malignancy around all industrial complexes that include major oil refineries in Great Britain. there was an apparent deficit of cases of Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, Keppel Street, London WClE 7HT,UK P Wilkinson B Thakrar P Walls M Landon C Grundy multiple myeloma near the refineries (p=0.04). Conclusion-There was no evidence of association between residence near oil refineries and leukaemias, or nonHodgkin's lymphoma. A weak positive association was found between risk of Hodgkin's disease and proximity to major petrochemical industry, and a negative association with multiple myeloma, which may be chance findings within the context of multiple statistical testing. (Occup Environ M e d 1999;56:577-580) &7 2, 4 7! SmallArea Health statistics Unit, Department of Epidemiology and Public Health, Imperial College School of Medicine, St hinry`s campus, Norfolk Place, London, W2 lPG, UK S Falconer P Elliorc Correspondence to: Dr Paul Wilkinson, Environmental Epidemiology Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WClE 7HT. Telephone 0044 171 927 2444; fax 0044 171 580 4524; email P.wi%nson@lshun.ac.uk Accepted 22 April 1999 Keywords: small area; lymphohaernatopoietic malignancy; oil refineries Oil refining and other petrochemical processes may lead to low level emissions of compounds with possible carcinogenic potential including benzene' and 1,3-butadiene.' A cohort study in the United Kingdom of mortality among oil refinery workers, who would be expected to have higher exposures to these compounds than residential populations, has not shown an excess risk of leukaemia or other haematopoietic malignancy.' Few studies of cancer risks among residential populations living near to petrochemical industry have been reported. A United States study found high rates of cancer of the lung, nasal cavity and sinuses, and skin among the resident male population in counties where the petroleum industry is most heavily concentrated, but no significant in- / 4-87, cL. 578 Wilkinson, Thakrar, Wds,et al Cell type 0- 2km 0 OIE (9% Cl) 0 - 7.5 km Stone 'r tests Unconditionalp Condirwnal p All ages All lymphatic and haematopoiet~c Multiple myeloma All leukaemias Lymphoid leukaemias Acute lymphatic leukaemia Chronic lymphatic leukaemia iMyeloid leukaemias Acute myeloid leukaemia Chronic myeloid leukaemia Hodgkin's lymphoma Non-Hodgkin's lymphoma Ages 0 to 14 All lymphatic and haernatopoietic A11 leukaemias Lymphoid leukaemias Myeloid leukaemias Hodgkin's lymphoma Non-Hodgiun's lymphoma 417 57 169 71 27 42 81 44 31 54t 137 25 19 16 2 2 4 0.95 (0.86 to 1.04) 0.74 (0.56 to 0.96) 1.06 (0.91 to 1.23) 1.00 (0.78 to 1.26) 1.09 (0.72 to 1.59) 0.99 (0.7 I to 1.33) 1.14 (0.91 to 1.42) 0.99 (0.72 to 1.33) 1.42 (0.97 to 2.02) 1.01 (0.76 IO 1.32) 0.92 (0.77 to 1.08) 0.99 (0.64 to 1.46) 1.00(0.60 to 1.56) 1.08 (0.62 to 1.76) 0.66 (0.08 to 2.38) 0.77 (0.09 to 2.80) 1.15 (0.31 to2.96) 3827 625 1446 597 197 364 666 420 207 443 1313 180 138 104 22 18 24 0.98 (0.95 to 1.01) 0.89 (0.82 to 0.96) I .01 (0.96 to 1.07) 0.94 (0.86 to 1.02) 0.95 (0.82 to 1.09) 0.91 (0.82 to 1.01) 1.05 (0.98 to 1.14) 1.05 (0.95 to 1.16) 1.06 (0.92 to 1.22) 0.97 (0.89 to 1.07) 1.OO (0.94 to 1.05) 0.93 (0.80 to 1.08) 0.94 (0.79 to 1.11) 0.91 (0.74 to 1.10) 0.93 (0.59 to 1.41) 0.89 (0.53 to 1.40) 0.91 (0.58 to 1.35) 0.64 0.04 0.58 0.24 0.24 0.31 0.31 0.57 0.25 0.05 0.99 0.37 0.73 0.60 0.74 0.07 0.28 0.51 0.52 0.78 0.55 0.17 0.77 0.38 0.50 0.71 0.02 0.77 0.12 0.38 0.38 t `Test did not converge because of small numbers +Includes 1 duplicateregistration. Conditional Stone's test for Hodgkm's disease remained significant (p=0.03) after its exclusion. Methods The study was carried out with the Small Area Health Statistics Unit postcoded health database. It examined cancer risk near all 11 oil refineries in Great Britain, operational by the early 1970s, which processed more than two million tonnes of crude oil in 1993. They were grouped into seven sites of one to three refineries (figure). Each was treated as an extended emissions source. The incidence of lymphohaematopoietic malignancy was examined in eight bands around refinery perimeters with outer limits at 0.5, 1,2,3,4.5,5.6,6.6, and 7.5 km.For descriptive purposes cancer incidence was also examined at 0-2 km and 0-7.5 km. The following diagnostic groups of the international classification of diseases 8th (ICD-8) and 9th (ICD-9) revision were studied for all ages and 0-14, 15-64, and 265 age groups: lymphatic and haematopoietic malignancy (ICD-8 200-207; ICD-9 200-208, 238.6); multiple myeloma (ICD-8 203; ICD-9 203, 238.6) (not separately examined for the 0-14 age group); non-Hodgkin's lymphoma (ICD-8 and 9 200,202); Hodgkin's lymphoma (ICD-8 and 9 201); all leukaemias (ICD-8 204-207; ICD-9 204-208); lymphoid leukaemia (ICD-8 and 9 204); and myeloid leukaemia (ICD-8 and 9 205). Acute and chronic subdivisions of the last two were examined for all ages. Cancer incidencewas studied for all years for which national cancer registration data were available: 1974-89 for England and Wales and Table 2 Observed (0) cmes and observedlexpected ranos (OlE) by distancefrom rejinery sites: all ages Outer radius of band (km) 0.5 1 2 3 4.5 5.6 6.6 7.5 Mulnple rn-veloma 0 OiE 4 0.46 8 0.58 45 0.82 55 0.83 103 0.79 128 0.87 147 1.03 I35 0.99 All leukaemias 0 OiE 14 0.74 25 0.84 130 1.17 137 0.99 275 1.02 316 1.07 279 0.96 270 0.99 Non-Hodgkin's Hodgkin 's lymphoma lymphoma 0 OIE 0 OiE ~~ ~~ ~~~ 14: 2.18 7 0.66 33 0.91 43 0.93 101 1.11 88 0.96 81 0.88 76 0.94 . 14 0.81 23 0.81 100 0.96 116 0.92 250 1.02 276 1.01 278 1.02 256 1.02 `Includes 1 duplicate registration, excluslon of which reduces the OiE ratio within 0.5 km to 2.02 1975-91 for Scotland. Population counts by band around each site were estimated from census small area statistics with the 1981 census for years up to and including 1986, and the 1991 census thereafter. Expected (E) numbers of cancer registrations were calculated by year and band from annual rates in Great Britain stratified by 5 year age group, sex, and quintde of the Carstairs deprivation index to adjust for possible socioeconomic confounding." Rates were further adjusted with disease specific standardised regional registration ratios." Because of the disparity in years for which cancer registration data were available, national rates for 1974 were computed from 1974 registrations for England and Wales and 1975registrations for Scotland; and 1990 and 1991 rates were computed from 1989 registrations for England and Wales and those for the relevant year in Scotland. Formal hypothesis tests were based on the maximum likelihood ratio test of Stone." This gives a test of decline in risk with distance against a null hypothesis of constant risk in all the bands. As Stone's unconditional test can be significant either where there is a decline in risk with distance or where the relative risk in at least one of the bands does not equal 1.0, we also carried out a conditional test. This specifically tests for decline in riskt4and consequently is the test of primary interest. Significance values were based on Monte Carlo methods using 999 simulations. Results from the seven industrial sites were combined by pooling the likelihood ratios from each one and comparing their sum with values obtained from simulated data. Results Results are presented here at all ages and for both sexes combined, and at ages 0-14 years. Age specific results, and results for each of the seven grouped industrial sites, are available on request. Over all refinery sites combined, there were 417 cases (439.9 expected) of lymphohaematopoeitic malignancy within 2 km of an oil refingry (observedlexpected (OiE) ratio il zi i b Ih 0.95,95% confidence interval (95% CI) 0.86 to Limitations to the interpretation of the study 1.04); and 3827 cases (3902.4 expected) within are imposed by several factors including lack of 7.5 km (OE 0.98, 95% CI 0.95 to 1.01, table exposure information. Measurements of vola- 1). None of the age and cell type specific tile organic compounds, including benzene, subgroups separately examined showed a signifi- near petrochemical plants have been made cant increase in risk within 0-2 km or 0-7.5 km. previo~sly2.8~F~or example, near the Baglan The conditional Stone's test for Hodgkin's Bay plantY9most of the measurements were < 5 lymphoma, uncorrected for multiple testing, ppb,28and with exposure to low concentrations suggested a decline in risk (p=0.02) with of ambient benzene there is great uncertainty distance from the sites, almost entirely due to an about the possible cancer risks." Other factors excess within 0.5 km (14* observed cases, 6.43 limiting interpretation include the small area expected, table 2). None of the conditional (ecological) nature of the study, uncertainties Stone's tests for other malignancies (including about population estimates for years between Hodglun's and non-Hodglun's lymphoma com- the census years and migration, and possible bined) suggested a decline in risk with distance unmeasured confounding factors. The results from the sites (tables 1 and 2). The significant cannot reliably be interpreted for specific cell (p=0.04) unconditional Stone's test for multiple types and age groups at indivdual sites, given myeloma at all ages reflected a deficit of these the reliance on routine data sources and the cancers from 0-7.5 km. fact that this would depend on multiple and non-independent analyses of subgroups, each Discussion Except for Hodgkin's disease, our study provides no evidence to suggest that populations living close to industrial complexes that include large oil refineries in Great Britain have an increased risk of lymphohaematopoeitic malignancy. This is consistent with the findings of broad scale studies of lymphatic and haemat- opoietic malignancy in areas containing petroleum works in the United States4l5and with occupational studies of oil refinery workers in the United Kingdom.' Although some studies have reported excesses of leukaemia in refinery workersYi6a recent overview of cohort studies from the United Kingdom and North America suggests no excess for cell type specific leukaemias," and oil refining probably conmbUtes little to environmentallevels of benzene and with limited statistical power. Overall the study is estimated to have had 80% power to detect an excess of around 11% in lymphohaematopoeitic malignancy at the 5% level within 2 km of refineries. We conclude that within the limitations of the small area approach based on routine data sources, there was no significant evidence of association between residence near to industrial sites containing major oil refineries and leukaemias or non-Hodgkin's lymphoma, and there was a negative association with multiple myeloma. Weak evidence of a positive association with Hodgkin's disease needs to be considered within the context of there being little, if any, other epidemiological evidence specifically relating Hodgkin's disease with the petrochemical industry. other petroleum products." The decline in risk with distance from oil This was a Small Area Health Statistics Unit project, supported by the Department of Health, the Department of the Environ- refineries for Hodgkin's disease must be viewed ment, Transport, and the Regions, the Health and Safety in the context of multiple testing and the gen- Executive, Scottish Office Home and Health Department, Welsh Office, and Northern Ireland Department ofHealth and eral tendency of Hodgkin's disease to cluster.20 Social Services. The Small Area Health Statistics Unit is funded That benzene might be a contributoryfactor in by grants from the Department of Health, Department of the Environment, Transport and the Regions, Health and Safery Hodglun's disease has been hypothesised by Executive, Scottish Office Home and Health Department, Aksoy et dJza1nd there have been reported Welsh Office, and Northern Ireland Department of Health and Social Services. We thank the Office for National Statistics and associations of Hodgkin's disease with occupation in the chemical industry," with other the Information and Statistics Division of the Scottish Health Service who made available postcoded cancer data, and, with the help of regional cancer registries, assisted in data checking. occupations entailing handling of We are grateful to representatives of the petrochemical industry chemical^,^^*^ and with exposure to diesel and local inspectorates of pollution for providing up to date information on site activities and geographical boundaries. The However, none of these studies provides direct evidence relating to the petro- views expressed in this publication are those of the authors and not necessarily those of the funding departments. chemical industry. In our study, there was a twofold risk of Hodgkin's disease (14 cases) within 0.5 km of refineries, and it is this which accounts for the positive Stone's test. This is compatible with a highly localised environmental hazard, although the risk within 2 km was almost identical to the regional expectation, and it was fractionallybelow expectationwithin 7.5 km. There was an unexplained deficit of cases of multiple myeloma within 7.5 km. ~ *Checkson a sample of cases revealed a duplicate registration of Hodgkin's disease in the 15-64 year age-band within 0.5 km of a rehery. After exclusion of this duplicate the OiE ratio for Hodgkin's disease (all ages) within 2 km was 0.99 (95% CI 0.74-1.30); the conditional Stone's test remained significant (~~0.03). 1 Austin H, Delzell E, Cole P.Benzene and leukemia. A m .7 Epufemiol1988;127:419-39. 2 Landrigan PJ. Critical assessment of epidemiologic sNdies on the human carcinogenicity of 1,3-butadiene. Enoimn Health Penpecr 1990;86:143-8. 3 Rushton L. A 39-year follow-up of the UK oil refinery and distribution center studies: results for kidney cancer and leukaemia. E w i m n Health Pmpecr 1993;101(suppl 6):7784. 4 Blot WJ, Brinton LA, Fraumeni JF,ec al. Cancer mortality in US counties with petroleum industries. Science 1977;198: 51-3. 5 Gottlieb MS, Shear CL, Seale DB. Lung cancer mortality and residential proximity to industry. Ermimn Health Perspecr 1982;45:157-64 6 Knox EG, Gilman EA. Hazard proximities of childhood cancers in Great Bntain 1953-80.7 Eptdemiol Community Healrh 1997;Sl: 151-9. 7 Bithell JF, Draper GJ. Apparent association between benzene and childhood leukaemia:methodological doubts concerning a report by b o x . 7 Eptdemior Community Health 1995;49:437-39. 8 Knox EG, Gilman EA. Migration patterns of children with cancer in Britain. .7Epidemwl Community Health 199832: 716-26. 580 Wilkinson, Thakrar, W&, et al 9 Sans S,Elliott P,Kleinschmidt I, et a[. Cancer incidence and mortality near the Baglan Bay petrochemical works, South Wales. Occup E n r i m Med 1995;52:217-24. 10 Lyons RA, Monaghan SP, Heaven M, et al. Incidence of Ieu- kaemia and lymphoma in young people in the vicinity of the petrochemicalplant at Baglan Bay, South Wales, 197491. Occup Envinm Med 1995;52:225-28. 11 Dolk H, Mertens B, Kleinschmidt I, er al. A standardisation approach to the control of socio-economic confoundingin small area studies of environment and health. J Epdemwl Community Health 1995;49(suppl2):9-14. 12 Elliott P, Shaddick G, Kleinschmidt I, et al. Cancer incidence near municipal solid waste incinerators in Great Britain. E r 3 Cancer 1996;73:702-IO. 13 Bithell JF,Stone RA. On statistical methods for analysing the geographical distribution of cancer cases near nuclear installations. JEpulemwl CommunityHealth 1989;43:79-85. 14 Shaddick G. Elliott P. Use of Stone's method in studies of disease risk around point sources of environmental pollution. Star Med 1996;15:1927-34. 15 Heary CD, Ury H, Siegelaub A, er al. Lack of association between cancer incidence and residence near petrochemi- cal indusuy in the San Fransisco Bay area. J Narl Cancer Inn 1980;64:1295-9. 16 McGraw DS, Joyner RE, Cole P. Excess leukaemia in a refinery population.3Occup Med 1985;27:220-2. 17 Wongsrichanalai C, Delzell E, Cole P. Mortality from leukaemia and other diseases among workers at a petroleum refinery.JOccupMed 1989;31:10611. 18 Wong 0, Raabe GK. Cell-type-specific leukaemia analyses in a combined cohort of more than 208 OOO petroleum workers in the United States and the United Kingdom, 1937-89. Re& Tmicol Phannacol 1995;21:307-21. 19 Wallace LA. Major sources of benzene exposure. Envimn Healrh Perspect 1989;82:165-89. 20 Alexander FE, Williams J, McKinney PA, et al.A specialist leukaemidymphoma registry in the UK. Part 2: clustering of Hodgkin's disease.E r J Cancer 1989;60:948-52. 21 Aksoy M,Erdem S, Din~olK, er al. Chronic exposure to benzene as a possible contributory factor in Hodgh's disease. Blur 1974;28:293-8. 22 La Vecchia C, Negri E, D'Avanzo D, er al. Occupation and lymphoid neoplasms.ErJ Cancer 1989;69:385-8. 23 Olsson H, Brandt L. Occupational handling of chemicals preceding Hodgkm'sdisease in man. EM3 1979;2:580-1. 24 Olin R. Leukaemia and Hodgb's disease among Swedish chemistry graduates. Lancet 1976;ii:916. 25 Garland FC, Gorham ED, Garland CF. Hodgkin's disease in the US Navy, IntJ EpidemwI1987;16:367-72. 26 Boffetta P, Stellman SD, Garfinkel L. Diesel exhaust exposure and mortality among males in the American Cancer Society Prospective Study. A m 3 IndMed 1988;14:403-15. 27 Wong 0. Bailey WJ.Cancer incidence and community exposure to air emissions from petroleum and chemical plants in Conua Costa County, California:a critical epidemiological assessment.3 Envirm Health 1993;56:11-17. 28 Bailey J, Campbell G, Kibblewhite M, et al. Air quality audit. BP Chemicals, Baglan Bay. LR 905 (APICA). Hertfordshire: Warren Spring Laboratory, 1992. (ISBN0 85624 765 0.) 29 United States Environmental Protection Agenct. Carcinogenic effects of benzene: an updare. Prepared by the National Center for Environmental Health, O&e of Research and Deuelopmenr. Washington, DC: USEPA, 1998. (EPA/600/ P-97/001F.) 1 Correspondence and editorials Occupational and Environmental Medicine welcomes correspondence relating to any of the material appearing in the journal. Results from preliminary or small scale studies may also be published in the correspondence column if this seems appropriate. Letters should be not more than 500 words in length and contain a minimum of references. Tables and figures should be kept to an absolute minimum. Letters are accepted on the understanding that they be subject to editorial revision and shortening. The journal also publishes editorials which are normally specially commissioned. The Editor welcomes suggestions regarding suitable topics; those wishing to submit an editorial, however, should do so only after discussion with the Editor. II f 598 Occup Environ Med 1999;56:598-601 Biological monitoring to assess exposure from use of isocyanates in motor vehicle repair N R Williams, K Jones, J Cocker Health and Safety Executive, 35 Dale End,Birmingham, B4 7NP, UK N R Williams Health and Safety Laboratory, Broad Lane, Sheffield, S3 7HQ, UK K Jones J Cocker Correspondence to: Dr N R Williams, Health and Safety Executive, 35 Dale End, Birmingham, 8 4 7NP, UK. Accepted 13 May 1999 Abstract Objectives-To develop a method for the measurement of a metabolite of hexam- ethylene diisocyanate (HDI), an isocy- anate, and use it to assess the exposure of sprayers employed in motor vehicle.repair shops. Methods-Urine samples were taken from sprayers wearing personal protective equipment and spraying in booths or with local exhaust ventilation, &om bystand- ers, and from unexposed subjects. Samples were analysed for a metabolite of HDI, hexamethylene diamine (HDA),by gas chromatogaphy-mass spectrometry (GC-MS). Results-HDA was detected in four sprayers and one bystander out of 22 workers. No HDA was detected in the urine of unexposed subjects. ConcZusions-Exposure to isocyanates still occurs despite'the use of personal protective equipment and the use of a booth or extracted space. Health surveillance is likely to be required to provide feedback on the adequacy of controls even if such precautions are used and to identify cases of early asthma. Biological monitoring can provide a useful additional tool to assess exposure and the adequacy of controls in this group of exposed workers. (Occup E m r o n Med 1999;56:598-601) Keywords: isocyanate; health surveillance; biological monitoring Isocyanates are highly reactive chemicals used in two component spray paints, adhesives, and in the manufacture of polyurethanes. They may also be released during cutting and welding of materials. Isocyanates are the commonest identified cause of occupational asthma in national' and regional' reporting schemes with the job title "spray painter" being the most commonly specified job.' Various methods have been reported for the measurement of isocyanates in air' and there has been considerable effort to improve the reliability of these methods. However, problems still remain. As control of exposure usually relies on air fed full face respirators, overalls, and gloves and the use of ventilated booths or ventilated spaces, air monitoring will give no indication of the actual exposure a worker receives. The efficacy of such measures to prevent exposure, subsequent sensitisation, and the development of occhpational asthma is not well established. Biological monitoring has the potential to measure a worker's body burden received by all routes of exposure (oral, inhalation, and dermal). Because of the frequency of occupational asthma in this group, workers using isocyanates are required to undergo health surveillance unless the risk assessment carried out under the Control of Substances Hazardous to Health Regulations 1994 (COSHH) draws the conclusion that there is little or no risk to their health from their workplace activities.` Several studies have reported that dermal contact with isocyanatescan induce respiratory symptoms in animals5 indicating that isocyanates may be dermally absorbed and biological monitoring has been used to study dermal absorption in humans.' Whether dermal absorption, if it occurs inhhumans, has any potential to lead to respiratory sensitisation is also unknown. Being able to estimate exposure to isocyanates through the development of a urinary assay for metabolites may be a useful way of assessing exposures by respiratory or dermal contact and the effectiveness of control measures. 1,6-HexamethyIene diisocyanate (OCN (CH,),NCO, (HDI)) is widely used in two component spray paints by the motor vehicle repair industry and is finding increasing use as a replacement for toluene diisocyanate.' Most HDI prepolymer based car paints contain a complex mixture of HDI monomers, polymers, HDI biuret, etc, but still contain small amounts of free HDI (0.5%-1%9). 1,6-Hexanediamine (HDA) has been detected in the urine (after hydrolysis) of car spray painters who use HDI based spray paints," and there have been several published methods for the analysis of HDA in Most of the published methods use acid hydrolysis with concentrated ( 36 M) hydrochloric acid.''-l4 Other studies" have used either alkaline or acid hydrolysis, reporting slightly better recoveries for the alkaline hydrolysis. We have studied both hydrolysis techniques but have persevered with the acid hydrolysis in line with most other researchers. Use of alkaline hydrolysis gave inconsistent results and variable, sometimes high, blank values. These problems were not encountered with acid hydrolysis. The free HDA is then derivatised and analysed, usually, by gas chromatography-mass spectrometry (GC- MS). Materials and methods SAMPLING STRATEGY Consent for the study was obtained from the Health and Safety Executive (HSE) research ethics committee. Five motor vehicle repair shops that use iso- FULL STUDY cyanate paints were identified by an inspector The full study involving five companies was from the HSE through routine visiting and carried out in the same way except that because local knowledge. They consisted of two na- of the nature of the activities only the firm in tional dealerships that employ full time paint the pilot study provided samples from bystand- sprayers and three smaller garages (two of ers. Samples were collected and despatched in which had a single owner who carried out a similar way, frozen and sent blind to the labo- intermittent spraying, and the third, which had ratory for analysis. employees who sprayed and did other work). All followed published HSE guidance in wear- ANALYTICAL PROCEDURE ing full face air fed helmets and spraying either 1,bHexanediamine (98%), 1,7-heptanedia- in a ventilated bootht6or in a controlled spray mine (HpDA), heptafluorobutyric anhydride t space." (HFBA), and hydrochloric acid were obtained . It was possible to recruit both subjects from Aldrich. Diethyl ether and toluene (high (sprayers) and controls at the same place of performance liquid chromatography (HPLC) work at three of the sites, at a fourth the unex- grade) were purchased from Rathburns. So- posed person was a member of HSE staff dium hydroxide solution (1 0 M) and sulphuric because the sprayer worked alone.At one of the acid were purchased from Fisher. Water was larger dealerships a group of people working purified through a Millipore Milli-Q system. just outside the spraying area were also An aliquot of HpDA solution (internal stand- recruited to act as bystanders. All participants ard) was added to the sample urine (2 ml). were provided with an information sheet and Concentrated sulphuric acid (200 pl) was then signed a consent form after discussing the added. Samples were mixed thoroughly and study with one of the researchers (NRW). All then hydrolysed at 100C for 90 minutes. sprayed alone in the booth or ventilated space. Samples were then cooled and 2 mi sodium Due to the reportedly short half life (about hydroxide solution (10 M) was added. 1.5 hours") of HDA in urine, samples were After hydrolysis, 4 ml diethyl ether was collected before, during, and after the shift to added and the samples were extracted for 20 detect any short term exposure as well as an minutes, then centrifuged. About 3 ml of the estimate of the 8 hour time weighted average organic layer was transferred and evaporated to exposure. i dryness under a stream of nitrogen. Samples were then resuspended in 500 pl toluene and derivatised with 50 pl HFBA at 55C for 1 PILOT STUDY hour. Samples were then cooled, evaporated to A pilot study was undertaken in 1996 at one of dryness, and reconstituted in 100 pl toluene. the larger dealerships. The samples were analysed by GC-MS with Subjects provided urine samples on the day negative ion chemical ionisation (with methane of spraying before starting spray work, at in the as reagent gas). Samples (1 pl, splitless) were middle of the shift, and at the end of the day. injected (at 350C) onto a BP-5, 25 mx0.32 They recorded the amount of time spent mm internal diameter fused silica capillary col- spraying and the personal protective equip- umn (1 pm film thickness). The oven tempera- ment worn. Bystanders and controls also ture ramped from 150C (initially held for 1 provided three samples at the same times as the minute) to 280'C at 10"C/min where it was sprayers. held for 1.5 minutes. The interface tempera- The samples were labelled with the person's ture was 280C and the source was held at name and either the time or the number of the 200C. With selected ion monitoring, ion mlz sample but without the details as to whether 449 was monitored for HDA and m/z 462 for they were sprayer, unexposed, or bystander. HpDA. Samples were frozen for despatch and sent by Concentrations of HDA in urine were courier to the Health and Safety Laboratory in corrected for creatinine concentration accord- Sheffield for analysis. ing to the method of Jaffe er a1.I' Table 1 Dernograplry and control measures of companies involved Company number 1 5 Demography and control measures ~~ A large dealership with full time sprayers working on large vehicles in a designated area with local exhaust ventilation. Sprayers wore full face air fed respiratory protective equipment (WE),disposable overalls (worn with the hood down exposing the head), and latex gloves. This was a small garage with a single ownerisprayer carrying out mtermittent spraying of cars in a homemade ventilated booth. While spraying he wore a full face air fed helmet with disposable overalls (hood down) and no gloves. A large dealership with full time sprayers carrying out car spraying in a Ventilated booth with full face air fed W E , disposable overalls, and latex gloves. A small garage with an ownerisprayer carrying out car sprayxng intermittently. He wore a full face air fed mask and spraved UI a ventilated booth wearing disposable overalls (with the hood down) and latex gloves. This was a small garage with several designated spravers who sprayed cars in a ventilated booth wearing full face air fed W E and wearing disposable overalls and latex gloves. Results ANALYTICAL METHOD The method was shown to be linear (defined as a least squares regression coefficient >0.99) over the concentration range 0-500 nmoVl with a detection limit of 5 nmoV1. The within and between assay coefficients of variation for the method were 4% and 10% respectively (n=lO). PILOT STUDY No HDA was detected in any sample taken before the shift nor in any of the samples provided by bystanders or unexposed subjects. Low levels of HDA (1 and 2 nmoV1) were detected in samples from the middle and the end of the shift respectively for two of the sprayers. 600 Williams,Jones, Cocker Table 2 ~ Urinary results and spray time detailsfor the full study HLM concenrratwn (pmollmol creatinine) Company Worker Sample Sample Sample Sex 1 2 3 Spraying time (?I) between samples Samples Sumples 1-2 2-3 1 Sprayer M 11.00 10.00 4.00 2.00 0.00 Sprayer M ND ND ND 3.00 2.50 Sprayer M ND ND ND 3.00 2.50 Bystander M ND ND ND NA Bystander M ND ND ND NA Bystander M ND 6.00 12.00 NA Unexposed M ND ND ND NA Unexposed F ND - ND NA Unexposed M ND ND ND NA Unexposed F ND ND ND NA 2 Sprayer MNDNDND 1.25 0.00 Unexposed F ND - - NA 3 Sprayer M ND ND 1.00 0.25 0.10 Sprayer MNDNDND 0.30 0.60 Sprayer M ND ND ND 0.75 0.30 Unexposed MNDNDND NA 4 Sprayer M ND ND ND 0.75 0.50 Unexposed FNDNDND NA 5 Sprayer' M ND ND ND NA Sprayer M 1.00 ND ND 3.00 1.00 Sprayer M ND 3.00 1.00 3.00 1.00 Unexposedsubjea M ND ND ND NA *No spraying undertaken on the sampling day. -No sample provided; ND=none detected; NA=not applicable. FULL STUDY A total of 11 sprayers, three bystanders, and eight unexposed subjects were identified from the five participating companies. Samples were collected from all participants except for one unexposed subject who only provided two samples. The companies are described in table 1 and the urinary results are presented in table 2. None of the unexposed subjects had detectable concentrations of HDA in any of the three samples. One bystander (company 1) had notable concentrations of HDA in samples from the middle and at the end of the shift. Four sprayers had concentrations of HDA. One in particular (company 1) had high concentrations in the samples taken before and in the middle of the shift, which fell by the end of the shift. Discussion ANALYTICAL METHOD The method was shown to be linear over the concentration range studied. The coefficient of variation between assays (10%) and detection limit (5 nmol/l) were comparable with other recently reported methods." l4 PILOT STUDY The pilot study showed that sprayers could receive low internal exposure to isocyanates despite wearing respiratory protective equipment and spraying in a ventilated area. It also showed that it is possible to spray isocyanates while using control measures without receiving a detectable exposure. In the pilot study none of the bystanders or unexposed subjects had any detectable concentrations of HDA. FULL STUDY The full study again showed that workers can absorb isocyanates during spraying operations despite wearing full face hir fed masks and using ventilated spray booths or extracted work areas. Four sprayers had detectable concentrations of HDA (two had detectable concentrations in their sample before the shift) and one bystander had detectable concentrations of HDA. Workers for small companies are not necessarily more at risk of exposure. Indeed, in this study the companies with only one sprayer (companies 2 and 4) had no detectable concentrations of HDA in urine. Also, the duration of spraying did not affect the likeli- hood of exposure (table 2, sprayers at company 1). Biological monitoring does not determine the route of exposure-this could have been oral, dermal, or through inhalation although inhalation is the most likely route given the lack of published evidence for the absorption of isocyanates through the skin. During feedback with participants it became clear that exposure may occur when the sprayers lift their visors to inspect the quality of their work-this was reported at several garages. Field work done by the HSE has shown that isocyanate concentrations in booths and spray spaces may remain considerable (>20 &m3 (in excess of the maximum exposure limit)) for some time after spraying has stopped. Clearance times of up to 5 minutes for bespoke booths and up to 30 minutes for spray spaces have been recorded. If sprayers and other workers are not aware of the importance of clearance times they may inadvertently expose themselves sometime after spraying has finished. Similarly baking during the spray bake cycle can drive off isocyanate monomer creating considerable concen- trations in the vapour within the booth or spray space during baking and for sometime after. In the case of company 1, the work involved spraying large lorries rather than cars and this involved some spraying above head height-an activity which may be relevant for the fit of respiratory protective equipment and subsequent detection of the metabolites in the urine of sprayers at this company. The short initial half life of HDA means that the urine sample only represents exposure over the past 2-4 hours. It is therefore possible that part time sprayers may have been exposed but that metabolite concentrations were too low and too quickly excreted to show that an exposure had occurred. However, it has also been r e p ~ r t e dt'h~at low amounts can be detected in urine 15-20 hours after an oral dose of HDA. There is therefore the possibility of accumulation of HDA in the body during the working week. Recent work on other isocyanates' has also shown that other adducts of isocyanate in plasma may break down over time causing excretion of the diamine metabolites for an extended period. Concentrations of toluene diamine (up to 5 pmol/mol creatinine) could be detected several days, even weeks, after exposure had ended. It is possible that this may also be true for HDA. However, the concentration found in the samples reported here is higher than previously reported for adduct breakdown, and the wash out found would be more consistent with the initial half life (1.5-2 hours) than adduct Further work is recommended to conduct a breakdown. It would therefore seem that the cross sectional indusmal survey and the possi- most likely explanation of the positive samples ble establishment of a biological monitoring before the shift would be an inadvertent short benchmark value to assess control of exposure. term high exposure before the sample was pro- vided. The sprayer at company 1 with positive concentrations of HDA had a possible exposure before the sample taken before the shift. Urinary concentrations of HDA were main- This study was supported by the Health and Safety Executive. We acknowledge the assistance of Mr Jon Harper, HM Inspector of Health and Safety, in the identification and recruitment of companies for the study. tained while spraying for 3 hours before the sample taken in the middle of the shift. Between providing that sample and the one taken after the shift no spraying was undertaken; this is reflected in the drop in the urinary HDA concentration (suggesting a half life consistent with that reported of about 2 hours). In the case of the bystander at company 1 discussion with the person showed that while not spraying himself he did think that he may have gone to speak to a sprayer at work. The concentrations of HDA detected in this study are similar to those found in other studie~.~l4 ' Brorson et all' reported a mean value of HDA after the shift of 20 pmol/mol creatinine after 7.5 hours exposure to 25 ~ g / m ' HDI whereas Maitre et all4measured a value of 12 pmol/mol creatinine for HDI exposure at the same concentration from a survey of workers. The study has shown that it is possible to 1 Meredith SK, McDonald JC. Work related respiratory disease in the united Kingdom 1989-92: report on the SWORD project. Occup Med 1994;44:183-9. 2 Siriruttanapruk S,Burge PS.Annual report of SHIELD in 1996 on behalf of the Midlands Thoracic Society Research Group. Circulated to participatingphysicians. Birmingham: MTSRG, 1996. 3 Lind P, Dalene M, Skarping G, et al. Toxicokinetics of 2,4and 2,6-toluenediamine in hydrolysed urine and plasma after occupational exposure to 2,4- and 2,6-toluene diisocyanate. Occup Em'nm Med 1996;53:94-9. 4 The Stationery Office. Control of substances hazardous to health regulations 1994: approved codes of practice. London: SO, 1994. 5 Karol MH, Hauth BA, Riley EJ, er al. Dermal contact with toluene diisocyanate produces respiratory tract hypersensitivity in guinea pigs. T e d Appl Pharmacol 1981;58:22130. 6 Rattray NJ, Botham PM, Hem PM, er al. Induction of respiratory hypersensitivityto MDI in guinea pigs. Influence of route of exposure. Toxicdog\. 1994;88:15-30. 7 Maitre A, Berode M, Stoklov M, et al. Human skin penetration study of hexamethylene diisocyanate (HDI) by using biological monitoring to hexamethylene diamine (HDA). International Symposium on Bio&gualMottiwring in Ompatwnal and Envinmmenral Healrh. 1996. ISBN 951-802136-8. 8 Redlich CA, Karol MH, Graham C, et al. Airway spray isocyanates without exposure to them if respiratory protective equipment and suitable extraction is provided but the results also suggest that exposure can occur despite the use of such precautions. Much effort has rightly gone into publicising the importance of the selection, fit, and maintenance of respiratory protective equipment; such effort will be wasted if consideration is not also given to the behaviour which workers adopt to do their job-that is, a human factors approach to prevention of ill health due to isocyanates is needed. isocyanate-adducts in asthma induced by exposure to 1,6hexamethylenediisocyanate.S c a d J @rk Emiron Health 1997;23:227-31. 9 Hazardous substances databank. Computerised database service. Washington, DC: HSDB, 1993. 10 Rosenberg C, Savolainen H. Determination in urine of diisocyanate-derived amines from occupational exposure by gas chromatography-mass fragmentography. Analyst 1 9 8 6 ; l l l : 1069-7 1. 11 Dalene M, Skarping G, Timerberg H. Biological monitor- ing of hexamethylene diisocyanate by determination of 1,&hexamethylene diamine as the nifluoroethyl chloroformate derivative using capillary gas chromatography with thermionic and selective ion monitoring. J C h m a m g r B Bwmed ApPr 1.994;656:319-28. 12 Dalene M, Skarping G, Brorson T. Chromatographicdetermination of amines in biological fluids with special reference to the biological monitoring of isocyanates and amines. Tv.Determinationof 1,6-hexamethylenediaminein Conclusion This study has shown that exposure to isocyanates can occur despite the use of full respiratory protective equipment and ventilated booths. It has also shown that urinary metabolites (HDA) can be used to identify these exposures and to assess the adequacy of controls. The evidence of exposure despite precautions highlights the continued need to improve control procedures, to provide health surveillance to check the efficacy of controls, and to provide information and instruction and training for workers in motor vehicle repair shops. human urine using capillary gas chromatography and selective ion monitoring. 3Chmmamgr 1990;516405-13. 13 Brorson T, Skarping G, Nielsen J. Biological monitoring of isocyanates and related amines. II Test chamber exposure of humans exposed to 1,6-hexamethylene diisocyanate (HDI). Inr Arch Ocmp E n v i m Healrh 1990;62:385-9. 14 b i n e A, Berode M, Perdix A, er al. Urinary hexane diamine as an indicator of occupational exposure to hexamethylenediisocyanate. lnr Arch Occup E n v i m Health 1996;69:65-8. 15 Skarping G, Dalene M, T i e r b e r g H. Biological monitoring of hexamethylene and isophorone isocyanate by the determination of hexamethylene and iscphorone diamine in hydrolysed urine using liquid chromatography and mass specuomeny. Analysr 1994;119:205 1-5. 16 Health and Safety in Engineering Mrkshops (HS(G)129). London: Health and Safety Executive, 1997. 17 Health and Safety in M o m Vehicle Repair (HS(G)67). London: Health and Safety Ewecutive, 1991. I 8 Jaffe M. Measurement of creatinine using picric acid. Z PhyswlChem 1886;10391-400.