(a) andoccup environ med: first published as 10.1136/oem.51.8.513 on 1 august 1994. downloaded from...

Occupational and Environmental Medicine 1994;51:513-518

Exposure of iron foundry workers to polycyclicaromatic hydrocarbons: benzo (a)pyrene-albuminadducts and 1-hydroxypyrene as biomarkers forexposure

0 Omland, D Sherson, A M Hansen, T Sigsgaard, H Autrup, E Overgaard

Department ofOccupationalMedicine, AarhusUniversity Hospital,Denmark0 OmlandDepartment ofOccupationalMedicine, VejleCounty Hospital,DenmarkD ShersonE OvergaardDepartment ofChemistry andBiochemistry,National Institute ofOccupational Health,DenmarkA M HansenSteno Institute ofPublic Health,Department ofEnvironmental andOccupationalMedicine, UniversityofAarhus, Denmark0 OmlandT SigsgaardH AutrupCorrespondence to:Dr 0 Omland, Departmentof Environmental andOccupational Medicine,University of Aarhus,Universitetsparken, Building180, DK-8000 Aarhus C,Denmark.

Accepted 19 April 1994

AbstractExposure to polycyclic aromatic hydro-carbons (PAHs) in foundry workers hasbeen evaluated by determination ofbenzo(a)pyrene-serum albumin adductsand urinary 1-hydroxypyrene. Benzo-(a)pyrene binding to albumin and1-hydroxypyrene were quantitativelymeasured by enzyme linked immuno-sorbent assay (ELISA) and reverse phasehigh performance liquid chromatogra-phy (HPLC), respectively. 70 malefoundry workers and 68 matched controlswere investigated. High and low exposuregroups were defined from breathing zonehygienic samples, consisting of 16 PAHcompounds in particulate and gaseousphase. Mean total PAH was 10-40 puglmin the breathing zone, and mean dustadsorbed PAH was 0*15 4ug/m.3 All car-cinogenic PAH was adsorbed to dust.Median benzo(a)pyrene-albumin adductconcentrations (10-90% percentiles) weresimilar in foundry workers (smokers 0 55(0.27-1.00) and non-smokers 0 58 (0.17-1-15)) pmol/mg albumin and age matchedcontrols (smokers 0 57 (0.16-1.45) andnon-smokers 0*70 (0.19-1.55) pmol/mgalbumin). Median 1-hydroxypyrene con-centrations were significantly higher (P <0.0001) in smoking and non-smokingfoundry workers (0.022 (0.006-0.075) and0*027 (0-006-0.164)) ,umollmol creatininethan in smoking and non-smoking con-trols (0 (0-0 022) and 0 (0-04010) ,fmollmolcreatinine). Dose-response relationsbetween total PAH, pyrene, carcinogenicPAHs, and 1-hydroxypyrene for smokers,and polycyclic aromatic hydrocarbonsadsorbed to dust for non-smokers aresuggested. Exposure to PAHs adsorbedto dust showed an additive effect. Therewas no correlation between the con-centrations of 1-hydroxypyrene andbenzo(a)pyrene-albumin adducts. Thechange in 1-hydroxypyrene over a week-end was also studied. Friday morningmedian 1-hydroxypyrene concentrationswere significantly higher in both smokersand non-smokers (0-021 (0-0.075) and0-027 (0-006-0-164)) ,umolmol creatininethan Monday morning median concen-trations (0.007 (0-0.021) and 0-008(0-0-021) umol/mol creatinine). Smokingdid not affect the concentrations of1-hydroxypyrene or benzo(a)pyrene-

albumin adducts. These data suggest that1-hydroxypyrene is a sensitive biomarkerfor low dose PAH exposure. Exposure toPAHs may be aetiologically related toincreased risk of lung cancer in foundryworkers.

(Occup Environ Med 1994;51:513-518).

Increased risk of lung cancer has been shownto occur in several studies of iron foundryworkers.l1 Polycyclic aromatic hydrocarbons(PAHs) may be an aetiological factor.56Exposure to carcinogenic chemicals (forexample PAHs) can be evaluated by measur-ing covalent binding with macromoleculartargets, DNA, and proteins.7 As target tissueis not usually available several surrogate mate-rials have been studied-namely, white bloodcells,8 lymphocytes,9 and serum proteins.101'1-Hydroxypyrene, a urinary metabolite ofpyrene, has recently been used as to monitorPAH exposure in humans.'2-15

In iron foundries PAHs are present in boththe gaseous phase and adsorbed on to dust.'6Benzo(a)pyrene adsorbed to particles in thelungs may augment carcinogenicity due toincreased long term retention.'7 18 These find-ings suggest that inhaled PAHs adsorbed toparticles covalently interact with macromole-cules to a greater extent than gaseous PAHs.'9The purpose of this study was to: (a) quan-

tify the concentrations of serum benzo-(a)pyrene-albumin adducts and urinary1-hydroxypyrene in healthy iron- foundryworkers, (b) investigate possible positive cor-relations between these biomarkers and PAHexposure variables (total PAHs, particulatePAHs, carcinogenic PAHs, and pyrene), (c)determine whether positive correlations existbetween the two biomarkers, and (d) compareFriday and Monday morning 1-hydroxy-pyrene concentrations.

Subjects and methodsSTUDY POPULATION AND CONTROL GROUPThe study population consisted of 70 maleiron foundry workers. They were employed inmelting, machine moulding, casting, or sandpreparation. All were occupationally exposedto PAHs. The foundry was located in a smallDanish town. The control group consisted of68 male blue collar workers from a drinkingwater supply plant near Copenhagen, all

513

on April 1, 2020 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.8.513 on 1 A

ugust 1994. Dow

nloaded from

http://oem.bmj.com/

Omland, Sherson, Hansen, Sigsgaard, Autrup, Overgaard

Table I Comparison ofcontrol and exposed groups

Control group(water supply Exposed groupworkers) (foundry workers)(n = 68) (n = 70)

Smokers (n = 42) (n = 45)Age (y) 43 40Age range (y) 24-66 24-63Pack-years 18 (11) 16 (13)Urinary creatinine

(mmol/1) 11-3 (4-3) 17-8 (6 9)

Non-smokers (n = 26) (n = 25)Age (y) 43 43Age range (y) 26-64 24-60Urinary creatinine

(mmoml) 13-1 (6-2) 19-6 (6-8)

All workers were men. Values are mean (SD) unless statedotherwise.

occupationally unexposed to PAHs. Thesegroups were matched for age and smokinghabits (table 1). Information about employ-ment, smoking, alcohol consumption, respira-tory and urogenital symptoms, andmedications including the use of coal tarsalves were obtained from questionnaires andchecked by personal interviews. Tobacco, cig-ars, and cheroots smoked were translated tocigarettes when calculating pack-years.20 Theresearch protocol was accepted by the localethics committee.

EXPOSURE GROUP AND AIR SAMPLINGFoundry workers were subdivided into lowand high PAH exposure groups based onbreathing zone measurements (n = 17): (1)total PAH: low <10.00 pg/iM,3; high >10 00pg/M3 (2) particulate PAH: low <021 g/M,'3high >0-21 pug/M3 (3) carcinogenic PAH: low<O 10 jg/M,3; high 0 10. pg/m3 and (4)pyrene: low <O0 IO pg/Mr,3 high 010 pig/M.3The sample preparation and steps of

reversed phase high performance liquidchromatography (HPLG) have previouslybeen described in detail21-23 and are only sum-marised here. Air samples (600-800 1) werecollected at a flow rate of 1 9 I/min during thework shift (about seven hours). Particulateswere collected on a 37 mm glass fibre filter(Millipore AP 4003-705) in a standard cas-sette of polystyrene, and gaseous PAH com-pounds on a PVC tube in series with astyrene-divinylbenzene adsorption tube. Theconcentration of 16 selected PAH com-pounds24 was measured by comparison withthe chromatogram of standard PAH prepara-tions. Total PAH concentration was calcu-lated as the sum of the particulate andgaseous concentrations. Total particulate con-centration was calculated as the sum of thecompounds attached to the glass fibre filter.The total, concentration of carcinogeniccompounds was calculated as the sum ofthe concentrations of benzo(a)anthracene,benzo(b)fluoranthene, benzo(k)fluoranthene,benzo(a)pyrene, dibenz(a,h)anthracene, andindeno(1_,2,3-cd)pyrene.2'BLOOD COLLECTIONBlood samples (10 ml) were collected fromexposed subjects on a Friday morning in April1989. The blood was allowed to coagulate

overnight at 4VC. The coagulum was removedafter centrifugation at 3000 rpm. The serumsamples were frozen at - 20'C until analysed.

ANALYSIS OF BENZO(a)PYRENE-ALBUMINADDUCTS IN BLOOD SAMPLESThe level of benzo(a)pyrene-albumin adductswas detected by competitive enzyme linkedimmunosorbent assay (ELISA) as previouslydescribed. 10 26 Serum (2 ml) was isolated fromthe blood samples by centrifugation, andalbumin was precipitated by addition of satu-rated ammonium sulphate (final concentra-tion 50%). After removal of globulins bycentrifugation, albumin was precipitated fromthe supernatant by acidification with aceticacid.Albumin was redissolved in 500 Al of 0 1iM

sodium phosphate buffer, pH 7 4, anddigested by pronase. The concentration ofalbumin was determined by Biuret reagentassay. The digest was applied to a prewet C 18Sepak column. The column was washed with5 ml of 5% methanol to remove small pep-tides and amino acids. The PAH metaboliteswere eluted with 80% methanol (5 ml) andthe eluate was evaporated to 500 p1 in avacuum.

COMPETITIVE ELISAPolystyrene 96 microwell plates (NUNCimmunoplates, Nunc, Roskilde, Denmark)were coated with 7,8-dihydroxy-9, 10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene (BPDE)modified dextran. The competitive antigenwas diluted 1:10 and 1:100 in phosphatebuffered saline (PBS), and 100pl/well (fivewells for each dilution) was added, followedby the primary antibody 8E11 (kindly pro-vided by Dr R Santella, Columbia University,NY, USA) diluted 1:250 000 in PBS (100pl/well). A series of different concentrations ofhydrolysed BPDE (0-1 fM to 1-0 pM, fourwells for each concentration) were included ineach plate. The level of modification was esti-mated from the standard curve with the1:1000 dilution of the competitive antigen,and was expressed as pmol BPDE/mg albu-min. The competitive assay has beendescribed in detail elsewhere.27

Antibody 8E11, which detects BPDEs,released by acid hydrolysis of proteins, alsorecognises benzo(a)pyrene diols, triols, andtetrols and cross reacts with several otherPAHs, with different affinities.101' Conse-quently results are expressed as benzo-(a)pyrene equivalents that correspond to theamount of BPDE necessary to give the samelevel of inhibition.

URINE COLLECTION AND DETERMINATION OF1-HYDROXYPYRENE AND CREATININEThe urine samples from foundry workers werecollected as the second urinary voids onFriday morning and the following Mondaymorning in April 1989, and frozen at - 20'Cuntil analysed. Urinary creatinine wasmeasured to standardise the results.28The sample preparation and chromato-

graphy have been previously described in

514



.bmj.com

/O

ccup Environ M


ugust 1994. Dow

nloaded from

http://oem.bmj.com/

Exposure of iron foundry workers to PAHs biomarkers for exposure55

detail.29 The standard plot was linear for theconcentration range of the standard curve.Urine samples (10-0 ml) were buffered with10-0 ml 0-2 N sodium acetate buffer (pH 5-0)and hydrolysed enzymatically with 200 p1fl-glucuronidase/sulphatase (26 400 U fl-glu-curonidase and 440 U sulphatase) for 20hours at 37-50G in a shaking water bath. Afterhydrolysis the sample was centrifuged for 10

mnts at 3000 rpm (928 g) at ambient tem-perature. Samples of the standard solutionand the centrifuged sample were applied to aBond Elute sample enrichment and purifica-tion cartridge packed with C-18 (Varian,Harbor City, CA, USA). The cartridge waspied with 2-5 ml acetonitrile, followed by

10 ml water, and the hydrolysed sample waspassed through the cartridge at a flow rate of2-5 ml/min. Subsequently, the cartridge waswashed with 5 ml 0-2 N sodium acetate buffer(pH 5-0). The retained solutes were elutedwith 1-5 ml acetonitrile. Samples of 25 p1 ofthe final extracts were injected into the HPLCcolumn.

STATISTICAL ANALYSISData were analysed with the statistical pack-age for social sciences (SPSS).30 Statisticalmethods comprised Student's t test and non-parametric tests for paired and non-paireddata.32 All tests were two tailed.

ResultsThe mean (SD) total PAH concentration was10-40 (4-04) pIg/in.3 The mean dust adsorbedPAH concentration was 0-15 (0-21) repre-senting 1-44% of the total PAH concentra-tion. Table 2 shows the mean concentrationsof the 16 standard PAHs in gaseous and par-ticulate forms. Four of the six carcinogenicPAHs were measurable in the dust adsorbedform only. Pyrene was found in both phases.

Figure 1 shows the median values for 1-hydroxypyrene concentrations from exposedworkers and controls. Smoking and non-smoking foundry workers had significantlyincreased 1-hydroxypyrene concentrationscompared with their control counterparts.The median values (10-90% percentiles) forthe four groups were 0-022 (0-006-0-075),0-027 (0-006-0-164), 0 (0-0.022), and 0

Table2 Mean (SD) concentrations of16 gaseous and particulate PAHsParticulate PAHs Gaseous PAHs Total PAHs

PAH ~~~~~~~~(ug/rn3) (pg/rn') (gr

Naphthalene 0 (0) 9-68 (3-85) 9-68 (3-85)Acenaphthylene 0 (0) 0 (0) 0 (0)Acenaphthene 0 (0) 0-03 (0-17) 0-03 (0-17)Fluorene 0-004 (0-01) 0-08 (0-05) 0-08 (0-05)Phenanthrene 0.01 (0-04) 0-31 (0-16) 0-32 (0-17)Anthracene 0-01 (0-02) 0-04 (0-03) 0-05 (0-03)Fluoranthene 0-02 (0-07) 0-11 (0-15) 0-13 (0-20)Pyrene 0-004 (0-01) 0-01 (0-02) 0-01 (0-03)Benzo(a)anthracene* 0-01 (0-02) 0 (0) 0-01 (0-02)Chrysene 0-02 (0-03) 0 (0) 0-02 (0-03)Benzo(b) fluoranthene* 0-003 (0-01) 0 (0) 0-003 (0-04)Benzo(k)* fluoranthene 0-02 (0-04) 0 (0) 0-02 (0-04)Benzo(a)pyrene* 0-02 (0-04) 0 (0) 0-02 (0-04)Dibenzo(a,h)* anthracene 0 (0) 0 (0) 0 (0)Benzo(g,h,l) perylene 0-05 (0-08) 0 (0) 0-05 (0-08)Indeno(1,2,3-cd)* pyrene 0 (0) 0 (0) 0 (0)Total (16 PAHs) 0-15 (0-21) 10-24 (4-03) 10-40 (4-04)

*Carciogenic PA~s. Seventeen measurements were taken.

Exposed smoker-Exposed non-smoker-

Control smokerControl non-smoker

0.001 0.01 0.1

HPU (jimoVmol creatinine)Figure 1 Median corrected 1-hydroxypyrene (HPU)concentrations with 10-90% percentiles forfoundryworkers and controls.P < 0.0001 smoking and non-smoking exposed workers vsmoking and non-smoking controls (Mann-Whitney Utest).

Exposed smokerExposed non-smoker I-H-

Control smoker ~j--Control non-smoker IIIHl

0.1 110BP-albumin adduct(pmoVmg albumin)

Figure 2 Median benzo (a)pyrene-albumin adductconcentrations with 10-90% percentiles forfoundryworkers and controls.P = 0-20 non-smoking exposed workers v non-smokingcontrols; P = 0-65 smoking exposed workers v smokingcontrols (Mann-Whitney U test).

(0-0-0 10) pmol/mol creatinine respectively.There was no significant difference in 1-hydroxypyrene concentration between smok-ers and non-smokers in the exposed or thecontrol groups (P exposed = 0-53, P unex-posed = 0-06).

Figure 2 illustrates the median benzo-(a)pyrene-albumin adduct concentrations inblood from foundry workers and controls.The adduct concentrations were not increasedin either smoking or non-smoking foundryworkers, compared with smoking and non-smoking controls (P smokers = 0-65, P non-smokers = 0-20). The median values for thefour groups were 0-55 (0-27-1-00), 0-58(0d17-1-15), 0-57 (0d16-1-45), and 0-70(0-19-1-55) pmol/mg albumin respectively.No difference in benzo(a)pyrene-albuminadduct concentrations was found betweensmokers and non-smokers in either theexposed or the control group (P exposed=0-99, P unexposed = 0-22).No correlations were found between any of

the PAH exposure variables (total PAHs,particulate PAHs, carcinogenic PAHs, andpyrene) and median benzo(a)pyrene-albuminadduct concentrations in the blood, either forsmokers or for non-smokers. Regressionanalyses showed no correlations between con-centration of 1-hydroxypyrene and albuminadducts (R2 = 5 x 10-5).

Figure 3 shows median 1-hydroxypyreneconcentrations according to total PAH expo-sure for smokers and non-smokers. Smoking

515



.bmj.com

/O

ccup Environ M


ugust 1994. Dow

nloaded from

http://oem.bmj.com/


High exposed smokerHigh exposed non-smoker

Low exposed smokerLow exposed non-smoker _

Control smokerControl non-smoker 1 , , 11A

0-001 0-01 0-1 1HPU (Amol/mol creatinine)

foundry workers in both low and high totalPAH exposure groups had significantly higherconcentrations than smoking controls. Themedian values for the three groups were 0-020(0-0-062), 0-022 (0-012-0-180), and 0(0-0-022) ,umol/mol creatinine respectively.Non-smoking foundry workers also had sig-nificantly higher 1-hydroxypyrene concentra-tions than non-smoking controls. Highexposure foundry workers had slightly lower1-hydroxypyrene concentrations than lowexposure workers. The median values were0-026 (0-008-0-083), 0-036 (0-006-0-174),and 0 (0-0-018) ,umol/mol creatinine respec-tively.

Figure 4 shows the median I-hydroxy-pyrene concentrations and exposure to parti-culate PAH for smokers and non-smokers.Non-smoking foundry workers with low orhigh exposures to dust adsorbed PAH had sig-nificantly higher concentrations of 1-hydroxy-pyrene than non-smoking controls. Themedian values for the three groups were 0-026(0-004-0-164), 0-036 (0-006-0-062), and 0(0-0-010) pumol/mol creatinine respectively.Both low and high exposure foundry workerswho smoked had significantly higher concen-trations of 1-hydroxypyrene than smokingcontrols. High exposure workers had slightlylower 1-hydroxypyrene concentrations thanlow exposure workers. The median valueswere 0-020 (0-0-057), 0-023 (0-001-0-178),and 0 (0-0-022) pmol/mol creatinine.The median 1-hydroxypyrene concentra-

tions according to exposure to carcinogenicPAHs and pyrene showed the same relation asfor total PAHs. Both smoking and non-smoking foundry workers had significantlyhigher concentrations of 1-hydroxypyrenethan their control counterparts (P low <0-0001, P high < 0-0001). For both exposurevariables smokers showed dose-response rela-tions whereas non-smokers did not.

Figure 5 shows Friday and Monday morn-ing median 1-hydroxypyrene concentrations.

High exposed smokerHigh exposed non-smoker

Low exposed smoker Jk I

Low exposed non-smoker _

Control smoker ,

Control non-smoker ..,,1

SmokerNon-smoker

SmokerNon-smoker

0-(

Friday /A

Monday J

H e |ill0-1

HPU (jgmol/I)Figure 5 Median 1-hydroxypyrene (HPU) Friday andMonday morning concentrations with 10-90% percentilesforfoundry workers.P < 0-01 smokers and non-smokers Friday HPU vsmokers and non-smokers Monday HPU (Wilcoxon ranksum Wtest).

Concentrations of 1-hydroxypyrene on Fridaymorning were significantly higher for bothsmokers and non-smokers than on Mondaymorning. The median Friday values were0-021 (0-0-075) and 0-027 (0-006-0-164)yumol/mol creatinine and Monday values were0-007 (0-0-021) and 0-008 (0-0-021)umol/mol creatinine.

DiscussionThe mean total PAH exposure in thisfoundry, 10-40,g/m3, is higher than the meanexposure recently described in anotherDanish foundry,'0 21 but lower than the meanexposure in German foundries (17-54,ug/m').'6 Similar exposure levels, (8-32,ug/m')have been found in roofers.33The mean concentration in our study of

both benzo(a)pyrene (0-0 1 ug/m3) and pyrene(0-01 ug/m3) was lower than previouslydescribed in another Danish foundry (0-04and 0-28 ,ug/m3)"0 and considerably lowerthan the mean concentration found in theGerman study (0-47 and 1-47 pug/m3). 16 Therange of concentrations in 10 Canadianfoundries for benzo(a)pyrene (0-004-8-15'ug/m3) and pyrene (up to 12-93 ug/m3)reflects a much higher exposure.'4 RecentFinnish studies have described a high expo-sure group with benzo(a)pyrene five timesgreater than ours (> 0-04 4ug/m3), thus indi-cating a substantially higher exposure inFinnish foundries."81The mean PAH adsorbed on dust was

1-44% of the measured total PAH concentra-tion. This ratio is considerably lower than inGerman foundries, where the PAH adsorbedon dust was 23-20% of the total PAH concen-tration. 16The reasons for these differences are uncer-

tain, but may reflect different sampling meth-ods, different binding materials, heterogenoushygienic conditions,'0 and seasonal differencesin air sampling.4As reported by others35 we found that all

carcinogenic PAHs were adsorbed on dust.This might be important considering the pro-longed retention of particulate PAHs found inanimal studies. 7-19

In our study, the level of albumin adductswas similar in the exposed and the control

Figure 3 Mediancorrected 1-hydroxypyrene(HPU) concentrationswith 10-90% percentilesfor controls andfoundryworkers according to lowand high exposure to totalPAH.P < 0-0001 low and highsmoking and non-smokingexposed workers comparedwith smoking and non-smoking controls (Mann-Whitney U test).

Figure 4 Mediancorrected 1-hydroxypyrene(HPU) concentrationswith 10-90% percentilesfor controls andfoundryworkers according to lowand high exposure to dustadsorbed PAHs.P < 0-0001 low and highsmoking and non-smokingexposed workers v smokingand non-smoking controls(Mann-Whitney U test).

IL,- - = s - - - ffi s E E - D . ^ .516

01 1



.bmj.com

/O

ccup Environ M


ugust 1994. Dow

nloaded from

http://oem.bmj.com/

Exposure of iron foundry workers to PAHs biomarkers for exposure

groups. Significantly raised protein adductconcentrations were found in Danish foundryworkers.10 Two Finnish studies8 11 foundhigher concentrations of adducts in occupa-tionally exposed foundry workers than in con-trols, but the increases were only slight in onestudy." In these three studies8101' the expo-sure to airborne benzo(a)pyrene was consider-ably higher than in our foundry and mayexplain why our present study did not finddifferences.

1-Hydroxypyrene in urine has been used asa biomarker in studying occupational expo-sure to PAHs in paving workers,"3 coke ovenworkers,'4 and foundry workers.'5 The resultsof these studies indicate that 1-hydroxypyrenemight be used not only as a marker forexposure to pyrene, but also to assess generalPAH exposure. Dutch studies of coke ovenworkers'436 found a more than 10-foldincreased concentration of 1-hydroxypyrenein exposed workers, and a twofold increase inbenzo(a)pyrene-DNA adduct. The presentstudy confirms that 1-hydroxypyrene is asensitive biomarker for exposure to PAHs.

There was no indication of a smoking effecton 1-hydroxypyrene concentrations in thisstudy although an effect has been reportedelsewhere.'0 1415 Others have shown similar 1-hydroxypyrene concentrations for smokersand non-smokers.'7 A recent study38 con-cluded that smoking increased the excretionof 1-hydroxypyrene, but it was not a strongdeterminant.Our data showed that foundry workers with

high exposure to PAHs had similar or onlyslightly greater 1-hydroxypyrene concentra-tions than foundry workers with low exposure.This suggests a limited exposure differencebetween the two exposure groups. It wasinteresting to note that the exposure variablePAH adsorbed to dust suggested a dose-response trend for non-smokers. This mightbe interpreted as an additional effect fromsimultaneous exposure to PAH and dust.Similar interpretations have been made fromresults with a combined PAH and silica expo-sure index.'0The concentration of benzo(a)pyrene-

albumin adducts in blood and urinary 1-hydroxypyrene in foundry workers did notshow any correlation. Possible high air pollu-tant PAH exposure in our control groupmight veil a real difference in PAH exposurebetween occupationally exposed and occupa-tionally unexposed workers, but this argu-ment does not explain why the twobiomarkers fail to mirror each other in thedata. The lack of correlation could be due todifferences in exposure profiles in the twogroups. Unsystematic misclassification tendsto reduce the risk estimate to unity, but aslong as the exposure variables were identicalfor the two biomarkers, this possible designerror does not present any real explanation.Differences between individuals in uptake,metabolism, and excretion might explain thelack of correlation between the two bio-markers. Another explanation could be thatthe PAH concentrations found in these

foundries is below the threshold for increasedbenzo(a)pyrene-adduct formation.The half life of 1-hydroxypyrene ranges

from six to 35 hours.'4 The Friday morningvalues for both smokers and non-smokerswere significantly higher than the correspond-ing Monday values for the same workers.These results correspond with a study of cokeoven workers'4 in whom the end shift valuesafter three days of exposure were higher thanthe preshift value.

Several potential confounders have beentaken into consideration. Only men werestudied. Age and duration of employmentwere not correlated with concentrations of thebiomarkers. There was no correlationbetween smoking and the concentration of thebiomarkers. None of the exposed workers orthe controls had been treated with coal tarsand no one had psoriasis. The study designdid not control for potentially important con-founding factors like exposure to airbornePAHs from the controls' working environ-ment and from general outdoor pollution.There was no reason to believe, however, thatblue collar workers from the drinking watersupply plant should be occupationallyexposed to PAH. Values for PAHs in theorder of ng/l have been measured in drinkingwater,25 so there is a theoretical possibility of aminimal exposure to PAHs from water. Thebenzo(a)pyrene exposure level at the presentfoundry is four to five times greater than thehighest concentration of benzo(a)pyrenedetected in Copenhagen (0009 ,ug/m3).39There are no data measuring the PAH con-tent in outdoor air from the town where thefoundry is situated. The community is busyand heavily industrialised. It is therefore likelythat the range of outdoor air pollution is simi-lar to that of Copenhagen. Possible differ-ences in PAH content from drinking waterand food2540 as well as liver function mighthave influenced the results to some degree,but this is likely to have been marginal.

We appreciate the financial support provided by the DanishHealth Service Fund.

1 Sherson D, Svane 0, Lynge E. Cancer incidence amongfoundry workers in Denmark. Arch Environ Health 1991;46:75-81.

2 Becher H, Jedrychowski W, Flak E, Gomola K,Wahrendorf J. Lung cancer, smoking, and employmentin foundries. Scand J Work Environ Health 1989;15:38-42.

3 Siverstein M, Maizlish N, Park R, Silverstein B, BrodskyL, Mirer F. Mortality among ferrous foundry workers.Am Jf Ind Med 1986;10:37-43.

4 Fletcher AC, Ades A. Lung cancer mortality in a cohort ofEnglish foundry workers. Scand J Work Environ Health1984;10:7-16.

5 Palmer WG, Scott WD. Lung cancer in ferrous foundryworkers: a review. Am Ind HygAssocJr 1981;42:329-40.

6 International Agency for Research on Cancer. Monographson the evaluation of the carcinogenic risk of chemicals tohumans. Vol 42. Silica and some silicates. Lyon: IARC,1987.

7 Ehrenberg L, Hiesche KD, Osterman-Golkar S,Wennberg I. Evaluation of genetic risks of alkylatingagents: tissue doses in mouse from air contaminatedwith ethylene oxide. Mutat Res 1974;24:83-103.

8 Perera FP, Hemminki K, Young TL, Brenner D, Kelly G,Santella RM. Detection of polycyclic aromatic hydro-carbon DNA adducts in white blood cells of foundryworkers. Cancer Res 1988;48:2288-9 1.

9 Haugen A, Becher G, Benestad C, Vahakangas K,Trivers GE, Newman MJ, Harris CC. Determination ofpolycyclic aromatic hydrocarbons in the urine, benzo-(a)pyrene diolepoxide-DNA adducts in lymphocyteDNA, and antibodies to the adducts sera from coke oven

517



.bmj.com

/O

ccup Environ M


ugust 1994. Dow

nloaded from

http://oem.bmj.com/


workers exposed to measured amounts of polycyclicaromatic hydrocarbons in the work atmosphere. CancerRes 1986;46:4178-83.

10 Sherson D, Sabro P, Sigsgaard T, Johansen T, Autrup H.Biological monitoring of foundry workers exposed topolycyclic aromatic hydrocarbons. Br J Ind Med 1990;47:448-53.

11 Lee BM, Baoyon Y, Herbert R, Hemminki K, Perera FP,Santella RM. Immunologic measurement of polycyclicaromatic hydrocarbon-albumin adducts in foundryworkers and roofers. Scand3 Work Environ Health 1991;17:190-4.

12 Jongeneelen FJ, Bos RP, Anzion RBM, Theuws JLG,Henderson PTh. Biological monitoring of polycyclicaromatic hydrocarbons; metabolites in urine. Scand 7Work Environ Health 1986;12:137-43.

13 Jongeneelen FJ, Scheepers PTJ, Groenendijk A, van AertsLAGJM, Anzion RBM, Bos RP, Veenstra SJ. Airborneconcentrations, skin contamination, and urinarymetabolite excretion of polycyclic aromatic hydrocar-bons among paving workers exposed to coal tar derivedroad tars. Am Ind HygAssocj 1988;12:600-7.

14 Jongeneelen FJ, van Leeuwen FE, Oosterink S, AnzionRBM, van der Loop F, Bos RP, van Veen HG. Ambientand biological monitoring of coke oven workers: deter-minants of the internal dose of polycyclic aromatichydrocarbons. Br] Ind Med 1990;47:454-61.

15 Sherson D, Sigsgaard T, Overgaard E, Loft S, PoulsenHE, Jongeneelen Fl. Interaction of smoking, uptake ofpolycyclic aromatic hydrocarbons, and cytochromeP450IA2 activity among foundry workers. Br ] Ind Med1992;49:197-202.

16 Knecht U, Elliehausen H-J, Woitowitz H-J. Gaseous andadsorbed PAH in an iron foundry. Br Y Ind Med 1986;43:834-38.

17 Creasia DA, Poggenburg JK, Nettesheim P. Elution ofbenzo(a)pyrene from carbon particles in the respiratorytract of mice. ] Toxicol Environ Health 1976;1:967-75.

18 Bond JA, Sun JD, Mitchell CE, Dutcher JS, Wolff RK,McClellan RO. Biological fate of inhaled organic com-pounds associated with particulate matter. In: Lee SD,Schneider T, Grant LD, Verker PJ, eds. Aerosols.Chelsea, MI: Lewis Publisher, 1986:579-92.

19 Wolff RK, Bond JA, Sun JD, Henderson RF, Harkema JR,Griffith WC, et al. Effects of adsorption ofbenzo(a)pyrene onto carbon black particles on levels ofDNA adduct in lungs of rats exposed by inhalation.ToxicolAppl Pharmacol 1989;97:289-99.

20 Silberschmid M, Sabroe S. Low dose asbestos exposure.Arhus: FADL, 1985;65. (In Danish, English summary.)

21 Hansen AM, Omland 0, Poulsen OM, Sherson D,Sigsgaard T, Christensen JM, Overgaard E. Correlationbetween work process related PAH exposure and urinarylevels of a-naphthol, fi-naphtylamine and 1-hydroxy-pyrene in iron foundry workers. Int Arch Occup EnvironHealth 1994;65:385-94.

22 Anderson K, Levin JO, Nielsson CA. Sampling and analy-sis of particulate and gaseous polycyclic aromatic hydro-carbons from tar sources in the working environment.Chemosphere 1983;12:197-207.

23 Hansen A M, Olsen IB, Poulsen OM, Holst E. Validationof high performance liquid chromatography/fluorescencedetection method for the simultaneous quantification of

fifteen polycyclic aromatic hydrocarbons. Ann OccupHyg 1991;35:603-11.

24 National Institute of Occupational Safety and Health.Manual of analytic methods. No 5506-1. Cincinnati:NIOSH, 1985.

25 International Agency for Research on Cancer. Monographson the evaluation of the carcinogenic risk of chemicals tohumans. Vols 32-35. Lyon: IARC, 1983-85.

26 Autrup JL, Schmidt J, Seremet T, Autrup H. Deter-mination of exposure to aflatoxins among Danish workersin animal-feed production through the analysis ofaflatoxin B adducts to serum albumin. Scand Y WorkEnviron Health 1991;17:436-40.

27 Autrup H, Seremet T, Sherson D. Quantitation of poly-cyclic aromatic hydrocarbons-serum protein and lym-phocyte DNA adducts in Danish foundry workers usingimmunoassays. In: Garner RC, Farmer FB, Steel GT,Wright AS, eds. Human carcinogenic exposure-biomonitoningand risk assessment. Oxford: IPL Press, 1991:207-13.

28 Bartels H, Bohmer M. Eine mikrometode zur kreatini-bestimmung. Clin Chim Acta 1971;32:81 -5.

29 Hansen A M, Poulsen OM, Hansen SH, Christensen JM.Determination of 1-hydroxypyrene in urine by highpressure liquid chromatography. Annals of Toxicology1993;17: 1-5.

30 Noruses MJ, ed. Statistical package for social sciences(SPSSIPS + V2 0). Base manual for the IBMPCIXTIA Tand PC12. Chicago: SPSS Inc, 1988.

31 Remington RD, Schork MA. Statistics with application tothe biological and health sciences. Englewood Cliffs:Prentice-Hall, 1970:310-25.

32 Schor SS. Fundamentals of biostatistics. New York: GPPutnam, 1986:174.

33 Herbert R, Marcus M, Wolff MS, Perera FP, Andrews L,Godbold JH, et al. Detection of adducts of deoxynrbo-nucleic acid in white blood cells of roofers by P-post-labeling. Scand]t Work Environ Health 1990;16:135-43.

34 Verma DK, Muir DCF, Cunliffe S, Julian JA, Vogt JH,Rosenfeld J. Polycyclic aromatic hydrocarbons inOntario foundry environments. Ann Occup Hyg 1982;25:17-25.

35 Bjorseth A, Becher G. PAH in work atmospheres: Occuranceand Determination. Florida: CRC-press, 1986: 117-39.

36 van Schooren FJ, van Leeuwen FE, Hillebrand MJX, deRijke M, Hart AAM, van Veen HG, et al. Determinationof benzo(a)pyrene-diol epoxide DNA adducts in whiteblood cell DNA from cokeoven workers: the impact ofsmoking. J Natl Cancer Inst 1990;82:27- 33.

37 Jongeneelen FJ, Anzion RBM, Theuws JLG, Bos RP.Urinary 1-hydroxypyrene levels in handling petroleumcoke. ] Toxicol Environ Health 1989;26:133-6.

38 Burgaz 5, Borm PJA, Jongeneelen FJ. Evaluation ofurinary excretion of 1-hydroxypyrene and thioethers inworkers exposed to bitumen fumes. Int Arch OccupEnviron Health 1992,63:397-401.

39 Ostenfeldt N. Mutagenic material in air pollution from thegreater Copenhagen area. Copenhagen: Danish Environ-mental Protection Agency, 1989. (Working documentNo 11.)

40 Birdbord K, French JG. Carcinogenic and mutagenic risksassociated with fossil fuels. In: Jones PW, FreudenthalR, eds. Carcinogenesis. Vol 3. New York: Raven Press,1978:451-63.

518



.bmj.com

/O

ccup Environ M


ugust 1994. Dow

nloaded from

http://oem.bmj.com/

(a) andoccup environ med: first published as 10.1136/oem.51.8.513 on 1 august 1994. downloaded from...

Documents