Urinary elimination of styrene in experimental and occupational exposure.

Urinary elimination of styrene in experimental and occupational exposure. Scand J Work Environ Health II (1985) 371- 379.Twentyhuman volunteers wereexposed to styrene vapor at 273-1 654!tmol/m' (28.4-172.3 mg/m') for a period of 1to 3h at rest (15cases)and during lightphysical exercise (5cases).Subsequently51workersoccupation allyexposed to styrenewerestudied duringa workweek (medianvalue 1 138!tmol/m', geometric standard deviation2.23).As expected, the relativeuptake averagedabout 65 lifo, and the ratio of the alveolarcon centration to the time-weighted average of the environmental concentration averagedabout 0.15. Both in the experimentally exposedsubjectsand in the occupationallyexposedworkersthe urinary styrenecon centration showeda linear relationship to the correspondingenvironmentaltime-weighted averagecon centration. The correlation coefficientsof the regressionlines ranged between0.88 (occupationallyex posedgroup)and morethan 0.93(experimentally exposedgroups).The regression coefficients wereclosely linked to the amount of styrene taken up and to the exposuretimes. The findingsshowthat the urinary styreneconcentration can be usedas an appropriate biologicalexposureindicator whosemeaningdiffers fromthat of othersuggested indices. Asan example, inoccupationally exposed subjectsperforming moderate work the urinary styreneconcentration correspondingto the time-weighted averageof the threshold limit value is 815 nmol/I, and the 95 0J0 lower confidencelimit (biologicalthreshold) is 740 nmol/1.

In industry styrene is a widely used solvent which can cause several toxic effects (28). It is a volatile compound entering the body mainly through the lungs or skin (8,31). General studies have been devoted to its uptake (1,37), distribution in body tissues (6,10), and elimination as urinary metabolites (mandelic and phenylglyoxylic acids) (11, 12,21 ,28). It is a substance easily soluble in the blood and biological tissues and therefore probably accumulates with each consecutive exposure (9,10).
For biological monitoring, determination can be made of the concentration of the parent (1,5,36) or metabolized compounds (1,11,28) in alveolar or expired air samples, in venous or arterial capillary blood samples, and in urine samples. Determination of its uptake during exposure has also been proposed as an exposure indicator (13).
More than 90 0J0 of the styrene taken up in the organism is excreted in urine as metabolites (28). Elimination via the lungs is less than 3 0J0 (6, 10, 13). Small amounts of styrene are eliminated as such by means of biological fluids (urine, sweat) (7).
In previous studies carried out during experimental and occupational exposure to certain solvents, the Reprint requests to: Dr G Pezzagno, II Sezione di Medicina del Lavoro, Via S Boezio 24, 1-27100 Pavia, Italy. relationship between the time-weighted environmental concentration and the urinary concentration of solvents as such (nonmetabolized amount) has been studied. The results suggest that the concentration of the solvent in a sample of urine (produced during the exposure period) could be considered an appropriate indicator of exposure from which the biological indices corresponding to the time-weighted average (TWA) of the threshold limit value (TL V) could be derived (17,23,29).
The main aim of our study was to verify the relationship between the urinary styrene concentration produced during a period of controlled exposure and the corresponding mean environmental concentration in the breathing zone.

Subjects
Experiments were performed in a laboratory on healthy volunteer subjects (stage 1) and directly on the work premises on groups of subjects occupationally exposed (stage 2).

Stage J
Twenty subjects, 14 males and 6 females (23 to 51 years of age), were exposed, in an open exposure chamber (16), to styrene at different concentrations but constant for each subject under study. The total group was divided into four five-subject subgroups. For each subgroup different exposure means were chosen as fol-  1 A) . The subjects went into the exposure chamber immed iately after emptying their bladder. At the end of th e chosen exposure period a urine sample was collected (for the precautions taken, see the Methods section) from which to measure the styrene concentration (C ll ) · Samples of chamber air, alveolar air, and mixedexpired air (the last sample being taken from the subjects of the first three subgroups only) were collected at regular intervals of 20 min. Pulmonary ventilation was measured simultaneously with the expired air sampling. Finally, immediately after the exposure, a venous blood sample was taken (from a brachial vein).
For all the subjects of the four subgro ups, the sampling of alveolar air continued for 10 min after the end of the exposure (at 30 and 150 s and then at 10 min) (figure 1 A).

Stage 2
The second stage of the study invol ved 51 male subjects (average age of the group 44 years) employed at a plastic boat factory (first factory) and in a plastic button factory (second factory). Shifts were 8 h dail y (0800-1600, with a half-hour lunch break) . The oc-cupational expo sure included, in addition to styrene, also acetone at a concentration of 2 80 1 JLmol/m 3 (med ian value) (geometric standard deviation = 1.95) (corresponding to 162.3 mg/rrr ').
Ur ine was collected from all the subjects at 0800, 1200, a nd 1600. A total of 101 sam ples was taken (figu re I B). In addition, during the workday, the enviro nmental concentration of styrene (breathing zone) was measured by means of personal pa ssive samplers (25).
The complete exposure period was monitored for each subject with four consecutive dosimeters and from calculations of the time-weighted individual value (Cj) from the four determinations.

Urine collection and styrene urinary concentrations
During stag es 1 and 2 the urine was collected from all the subj ects in the same manner in 120-ml glass vials (Supelco Vials) with airtight plu gs (without silicone). The vials were equipped with a valve (Minimert valve Sup elco) enabling a headspace gas sample of styrene to be withdra wn over th e urine after partial pre ssure equilibration . The volume of urine was mea sured by weight , start ing from the urine densit y.
The subj ects voided rapidly in areas known to be unpolluted. (Significant variations in C u do not occur if collection time is less than I min .) The vials were then stored at 37°C for 2 h and shaken periodically to speed the separation and partial pressure equilibrium of the styrene vapor between the urine and the headspace air over it. After at least 1 h a headspace air sample was analyzed by means of a gas chromatograph (HP 5880 A) plus a mass selective detector (HP 5970 A) . The analytical conditions for the gas chromatograph were as follows: cross-linked column 5 % phenyl methyl silicone (internal diameter 0.2 mm, length 25 m), column temperature 150°C, carrier gas helium with a flow rate of I mllmin, retention time of styrene 2 min 10 s, sampler volume (kept at 37°C) 0.5 m!' (The headspace air sample was transferred to the gas chromatograph sampler by means of a Hamilton gastight syringe.) The analytical conditions for the mass selective detector were: monitored ion 104, dwell time 50 rns, selected-ion-monitoring window time 0.1 amount mass units, electromultipler voltage 1 800 V. From the headspace styrene concentration it is easy to obtain the urinary concentration if the A urine/air value and the volume of the two phases [urine volume (V) and headspace air volume (Va)] are known.
[O~ce a balance between the two phases of the vial has been reached, the volume of dissolved solvent (x) is divided between Vu and Va proportionally to A (Ostwald partition coefficient) (x, and x 2 are the two volumes of solvent distributed to V u and Va' respec- is the measured amount of styrene in the headspace. The A urine/air of styrene, at 37°C, is 7.13 (17).J

Determination of venous blood concentration
Similar to what was described for the CiJ measurement, in the experimental exposure only, the heparinated venous blood sample (5 ml) was kept at 37°C in similar airtight vials (volume 13.5 ml) and shaken periodically for the determination of the venous blood concentration (C y ) . The analysis of the headspace samples and the blood concentration determination followed the same modalities described with regard to the CiJ determination. The A blood/air value used was 52 (32).

Alveolar and mixed-expiratory air sampling for the styrene concentration
In the experimental exposure only, the alveolar (C A ) and mixed-expired (C E) air samples were collected every 20 min during the exposure at the same time as the collection of environmental air from the chamber and the pulmonary ventilation measurement (V). Three additional air samples were collected at the end of the exposure (after 30 and 150 s and after 10 min) .
The open-circuit apparatus is schematized in figure  2. It had a oneway valve (through which the subject under study breathed), the inspiratory end of which was equipped with a pulmonary ventilation analyzer (ultrasonic spirometer LS 75 Bourns), and the expiratory end of which had a fourway tap allowing the expiratory flow to pass across an alveolar sampler and an expiratory air sampling bag, a plastic bag whose inner wall is an aluminum foil barrier (5 I of volume) which opens outside but is kept closed by means of a oneway valve letting the extra expired air out (Multilayer gas sampling bag, Calibrated Instruments). The bag was provided with a device to draw the air samples either into a gastight syringe to be analyzed or to the outside directly. The entire expiratory part of the circuit was placed in a metal container (kept at 37°C).
The alveolar sampler was a metal cylindrical device in which an alveolar air sample can be trapped and from which the sample can be partly taken out to be analyzed (22). Alternatively the alveolar sampling can be made by means of an airtight, heated syringe set up on the expiratory end of the circuit (figure 2).
The C A and C E measurements were made with the same gas chromatograph and mass selective detector   Table 2. Single values of the time-weighted average of the environmental concentration (Ci), the urinary concentration (C u)' (urine sampled immediately after exposure), the venous blood concentration (Cy) (blood withdrawn immediately before the end of exposure), and the uptake of styrene (U) during exposure. The last column shows the regression equation relating C u and Ci.
Subgroup and under the same analytical conditions used for the urine samples.

Time-weighted average of the environmental concentration measurement in the breathing zone
The activated charcoal of passive dosimeters, used in stage 2 only, were deabsorbed with 5 ml of carbon disulfide and then kept at room temperature (20°C) for 1 h and shaken periodically. Finally 0.5 jtl of the deabsorption liquid was injected into the gas chromatograph-mass selective detector unit.

Stage 1
The average concentration in the exposure chamber (Cj) ranged from 273 to 1 654 jtmollm 3 [the conversion factor from milligrams to micromoles: 1 mg = 9.6 jtmol (of styrene)] for the various subjects. During the exposure the chosen values for each subject were maintained at a relatively constant level (SD 23 jtmollm 3 ) .

374
The alveolar concentration value (C A) was also found to be consistent for all the subjects up to the third exposure hour with a CA:Cj ratio close to 0.15 (table 1).
The relative uptake (R) (measured from Cj and C E) R, was found to be 4.2 11m (in the first three subgroups) and 9.4l1m (in subgroup 4) (CV ± 7.1 and ± 21 %, respectively). The amounts absorbed (total uptake) during the exposure time ranged from 75 to 1 427 /Lmol (table 2).
The TWA environmental concentrations (Cj) and the urinary (C,,) and blood (C v ) concentrations obtained for the 20 subjects at the end of exposure are shown in table 2, together with the uptake (U) and exposure time. The C u values displayed a statistically significant linear relationship to the corresponding Cj (or U), and the C A or C, values were also statistically significant:

Stage 2
The same relationship between Ci and C u observed experimentally also existed for the group of subjects occupationally exposed to styrene vapor ( where Cj and Co are expressed in micro moles per cubic meter and namomoles per liter, respectively.

Discussion
The starting values of the relative uptake (R) and relative alveolar concentration (CA:Cj) of styrene depend on their A blood : air ratios; our results (R := 0.66, CA:q = 0.15) are consistent with those of previous investigations (1,3, 10, 15,37, 38). The Rand CA:Cj curve (1) during exposure depends on the possibility of the biological tissues to store solvent and on the rate of its biotransformation; our results , in accordance with those already published (I , 13) show a practically constant curve for R and CA:Cj throughout the entire exposure time (therefore at least up to the end of the third hour) in the range of the environmental concentration values chosen (273-1 654 JLmollm 3 or 28.4-172.2 mg/m").
The main aim of this report was, however, to determine if correlation relationships existed between styrene exposure standards (Cj) and the concentration of styrene in samples of urine produced throughout the exposure time (C u )' Previous attempts at analyzing styrene as such in urine were unsuccessful due to the inaccuracy of the , . The difference in the slope between the equations obtained after 1 and 2 h of exposure (0.11 against 0.16) was significantly higher than that between 2 and 3 h of exposure (0.16 against 0.17). This occurrence was probably due to a continuous increase in styrene concentration in arterial blood from the beginning of exposure up to I h or more later (6,10,37,38).
The urinary concentrations should be in fact proportional to those in the arterial blood, with a proportional coefficient corresponding to the A blood/ urine value, on the assumption that the exchange of the solvent between the arterial blood and urine occurs after simple partition in accordance with Henry's law.
In this case, once a constant arterial concentration value has been reached, the slope of the regression lines should be independent of exposure time. The small slope difference of the regression lines corresponding to 2 and 3 h of exposure (figure 5) should mean that an increment in the blood concentration is still occurring throughout the third hour of exposure or that a stable level of urinary concentration is not rapidly reached, a phenomenon similar to what happens to the alveolar (constant) and arterial concentration (increasing up to over 75 min of exposure) (37,38).
The greatest slope of the regression line in subgroup 4 (regression coefficient 0.26) was due to the work  load . Under such conditions, R values being equal, the amounts of styrene taken up are clearly larger (depending on pulmonary ventilation), and greater solvent concentrations are therefore reached in arterial blood (10), and consequently greater urinary concentrations as well.
Th e four regression lines star t pract ically from the origin of the Car tesian coordinates (figure 5) (intercepts o f the y axis between -2.6 and 21.7 nmo lll) because the subjects under study had never been exposed to styrene previously.
A close relationship between th e TWA environmental concentration in the breath ing zone (Cj) and the urinary concentration of styrene in urine samples collected after 4 h of exposure also existed for the group of occupationally exposed subjects (correlation coefficient 0.88), the regression coefficient being 0.32.
The regression line does not start from the origin of the Cartesian coordinates but intercepts the y axis at 154 nmolll ( figure 3 and 6); this finding indicates that at the beginning of the exposure period there was some styrene in the blood of the subjects under study, as expected and previously documented (9,10).
The urin ar y concentration (C u ) corresponding to the TLV-TWA (2 065 Jlmollm 3 ) is 815 nmolll ( figure  6). To propose a biological exposure limit -based on 4 h of occupational exposure and on conditions of a light work load -it is necessar y to consider the corresponding 95 % lower confidence limit of the regres- The biological indicators presently proposed to evaluate the occupational exposure to styrene are (I, 2,5,11,18,19): (i) the urinary concentration of mandelic and phenylglyoxylic acid, (ii) the styrene concentration in alveolar or mixed-expired air, and (iii) the styrene concentration of venous or arterial capillary blood. The proposed threshold values are 1 g (6.57 mol)!l for mandelic acid and! or 250 mg (1.66 mol)!1 for phenylglyoxylic acid in urine (where the urin e was collected at the end of the workshift), 40 ppb and 18 ppm of styrene in mixed-expired air (before and du ring the workshift, respectively), and 0.02 rng/ I and 0.55 mg! 1 for styrene in venous blood (before and during the workshift, respectively) (2).
The results obtained both experimenta lly in volunteers exposed in an exposure chamber and in groups J Complian ce exposure: 95 0J0 confidence based on measurements that a worker's exposure is below the standard; noncompliance exposure : 95 % confid ence that a worker's exposu re is above the sta ndard; possible overexposure: any exposur e which cannot be classified as comp liance or noncompliance exposur e.
Crr(nmor/I)  of subjects occupationally exposed enable us to include also the styrene urinary concentration in the already proposed biological exposu re indice. With regard to the meaning and the value used to ascribe to the various biological indicators proposed, it should be pointed out that the blood (C v ' CJ and air (C A, CE> measurements are instantaneous values, while the unchanged urinary styrene (C,,) and styrene urinary metabolites are weighted values. However they are not indicators of th e same exposure time. In fact the metabolic indicator phase depends on the rate of biotransformation. Its metabolites can also be classified as effect indicators . On the contrary the unchanged styrene urinary elimination dur ing exposure indicates a natu ral integrat ion over time of the rather fast partition between air and arte rial blood and between arterial blood and urine (the bladder serving as a collecting and mixing vessel). Besides, the situations of highest uptake of styrene (environmental concentration being equal) by subje cts working under different work loads are clearly indicated by the urinary concentration values. On the contrary it is well known that, mainly with highly soluble solvents (like styrene), increases in pulmonary ventilation do not bring about proportional increases in alveolar concentration (1,37). Thus C A values correlate well with q values (5) but not with the amounts of styrene taken up .
Finally it is to be pointed out that the unchanged styrene urinary concentration should not be affected either by other solvents interfering with each other in the biotransformation rat e or by the presence of common metabolites (38). Last, the method is simple, quickly performed, and noninvasive .
The exchange of styrene between renal arterial blood and urine could occur (similar to what is believed to take place at the alveolocapillary level) by means of a simple partition mechanism between these two fluids regulated by the A blood/urine value; in this case a pressure equilibrium should be reached with PA = P, = Pu (20) (and then C./C A = Ablood/air; C./C u = Ablood/urine). Such a hypothesis has already been advanced (existence of an alveolourinary inert gas partial pressure equilibrium) and employed in the field of pulmonary physiolo gy (4,24,30).
The mean value of C" in the group of subjects exposed for 3 h at rest (subgroup 3) was 145 nmol/I, and the mean value of CAin the same group was 103 JLmol/m 3 • The experimental value of the Ca:C A ratio corresponded to the in vitro A blood/urine value (37). Therefore the styrene arterial concentration (CJ (at equilibrium) is likely to be about C A • A, ie, in this case, 5 350 nmol/I. [No direct measurement of the styrene concentration in arterial blood was made because the introduction of a catheter into an artery was not included in the acceptance protocol of the volunteers and because reference data are already available (1,12,37,38).] On the assumption that the C.:C A ratio corresponds to the in vitro Ablood/urine value (A = 7.29) (17), the value of C u in the last parts of the urine, filtered towards the third hour of exposure, should be 735 nmol/l; the time-weighted value of C u ' of the entire urine sample filtered throughout the 3 h of exposure, was 145 nmol/I. Such a value is not in contrast with the hypothesis advanced. Moreover none of the C u values suggested an active mechanism of renal styrene elimination.