~c)Copyright 1987 by The Humana Press Inc. All rights of any nature, whatsoever, reserved. 0163-4984/87/1300-0241502.20
Biological Monitoring, by In Vivo XRF Measurements, of Occupational Exposure to Lead, Cadmium, and Mercury S T A F F A N S K E R F V 1 N G , ~'* J A N - O V E ANDREJS
CHRISTOFFERSSON, 2
SCHLITZ,' H A N S W E L I N D E R , ~ G U N N A R SP,a, NG, 3
LARS A H L G R E N , 2 A N D S O R E N M A T T S S O N 4
'Department of Occupational Medicine, University Hospital, S-221 85 Lund, Sweden; 2Department of Radiation Physics, Lund University, /rialto6 General Hospital, S-21401 ~alm6, Sweden; 3Department of Occupational Health, Kapellv~gen 2, S-5 72 O00skarshamn, Sweden; and 4Department of Radiation Physics, University of Gothenburg, S-413 45 Gothenburg, Sweden. ABSTRACT In vivo X-ray fluorescence (XRF) techniques were used for biological monitoring of lead, cadmium, and mercury. Lead accumulates in bone, the level of which may thus be used for monitoring of exposure. However, there was no close association between lead levels in bone and exposure time, partly because of differences in exposure patterns and partly, probably, because of variations in the toxicokinetics of lead. There are at least two separate bone lead compartments. The average over-all half-time is probably 5-10 yr. The finger bone level may be an index of the lead status of the total skeleton. In lead workers, the mobilization of bone lead causes an "internal" lead exposure and affects the blood lead level considerably. In cadmium workers, in vivo XRF is a sensitive and risk-free method for assessment of accumulation in kidney cortex, the critical tissue as to toxic effects; workers displayed increased levels. However, there was no clear association with duration and intensity of exposure, cadmium levels in urine, or microglobulinuria. Determinations of kidney cad*Author to whom all correspondence and reprint requests should be addressed. Biological Trace Elernent Research
241
VoL 13, 1987
Skerfving
242
et
aL
mium may add important information on the state of accumulation and, thus, risk of kidney damage. Workers exposed to elemental mercury vapor, as well as fishermen exposed to methyl mercury, had mercury levels in bone below the detection limit of the XRF method. Index Entries: Lead; bone; X-ray fluorescence; toxicokinetics, accumulation of; half-time; interindividual differences; cadmium; kidney; mercury.
INTRODUCTION The term "biological monitoring" usually denotes determination, in an index medium, of accumulated or excreted amounts of a compound. Instead of the compound itself, its metabolites may be analyzed. Alternatively, a biological effect caused by the exposure may be employed. Traditionally, the index media employed have been blood, urine, and, sometimes, hair. However, the modern methods for in vivo determination of elements in different organs by radioanalytical techniques have opened new possibilities for biological monitoring of exposure to metals. We report here some results regarding occupational exposure to lead, cadmium, and mercury. Biological monitoring may give information regarding exposure. This gives several advantages as compared to monitoring of exposure by air sampling: It is a true personal sampling, reflecting the inhaled amount of the compound. Also, it takes into consideration the fact that exposure may vary enormously at a certain air concentration, depending on the physical activity of the worker. Further, the problem of different absorption of particles of different size, often causing trouble in the evaluation of the health hazards associated with air levels, is avoided. Moreover, absorption through the gastrointestinal tract and the skin is included. Also, the biological monitoring may sometimes reflect a time-integrated exposure, thus indicating the intensity of the exposure far back in time, which may be important in evaluation of dose-effect and dose-response relationships. There are problems involved in the use of biological monitoring for evaluation of exposure. The toxicokinetics of the compound to be monitored must be known. Also, interindividual variations in kinetics may obscure the picture (1-3). In addition, biological monitoring may indicate risk of poisoning; nota bene if the rglationship between the level in the index medium and the threshold of the critical effect is known. Again, interindividual variations in metabolism and sensitivity must be known. We summarize here some data on the relationship between exposure to lead, cadmium, and mercury on the one hand, and concentrations of these metals in index media on the other, as well as studies on their toxicokinetics, including interindividual variation in metabolism Biological Trace Element Research
Vol. t3, t987
Biological~onitoring by In Vivo XRF
243
and the relationship between the metal levels in the index medium and toxic effect.
LEAD
Background Lead accumulates in the skeleton. On the basis of lead levels in different organs (4), it may be estimated that bone contains more than 90% of the body burden. Thus, concentrations of lead in bone might be useful for biological monitoring. The turnover of skeletal lead is usually assumed to be slow, and the level of lead in bone may therefore reflect the time-integrated exposure over a long time. But the evidence of a slow turnover is only circumstantial, and more information is needed. Considering the huge skeletal pool of lead, even a slow release may considerably affect the blood lead level, at present the main parameter used for biological monitoring of lead exposure (5). Also, a rapid mobilization could constitute a risk of poisoning.
Subjects and Methods In a cross-sectional study (6), 75 active lead workers (mainly smelter workers, storage battery workers, cast bronze founders, and demolition workers) were studied. In addition, we investigated 32 retired smelter workers. In a longitudinal study, 21 lead workers were followed after the end of exposure. Lead levels in finger bone were analyzed by an in vivo X-ray fluorescence (XRF) method, using 57Co as the radiation source (6-8). The detection limit is 20 ~g/g (three standard deviations above the background). The precision of the method is about 15%. In addition, lead levels were analyzed in bone biopsies obtained from the spinous process of a lumbar vertebra (9). Analysis of lead was performed by electrothermal atomic absorption after combined dry and wet ashing. The detection limit was about I ~g/g and the precision about 5%. Lead levels in blood and urine were determined by flame atomic absorption spectrometry (10). Time-integrated blood lead level was calculated in subjects who had been monitored for blood lead during their entire time (up to 11 yr) of occupational lead exposure (6). By use of the least-square method, monoexponential functions were fitted to the data on the accumulation and elimination of lead from the skeleton, and biexponential functions to the data on elimination of lead from the blood (11). Biological Trace Element Research
Vol. 13, 1987
244
Skerfving et aL
Results The median lead level in finger bone in the 75 active lead workers was 43 ~g/g; the corresponding figure in the 32 retired workers was 59 ~g/g (Fig. 1). There w a s a rise of lead levels in finger bone with increasing time of exposure. A function of the type, bone-Pb = A[1 - exp( - B x t), was fitted to the data (Fig. 1). A is a constant, B the elimination rate constant, and t the exposure time. The value of the elimination constant thus arrived at corresponds to a half-time for lead in finger bone of 6.3 yr. However, the interindividual variation was considerable; the function explained 45% of the total variance of lead levels in bone on time of exposure. In 43 subjects, the time-integrated blood lead level was compared to the lead levels in finger bone (6). There was a significant correlation, but the interindividual variation of bone lead at a certain integrated blood level was considerable. In 18 workers, w h o were temporarily removed from exposure because of high blood lead levels, determinations of lead concentration in finger bone were m a d e at the beginning as well as at the end of the exposure-free interval, which averaged 5 too. During that period, no systematic decrease of lead concentration in bone could be demonstrated. In 14 ex-lead workers, lead level in finger bone was followed after the end of occupational exposure (Fig. 2). There was a decrease in all but one. A monoexponential function was fitted to the data. The average half-time was 7 yr. In 21 ex-lead workers, blood lead level was followed for up to 13 yr after the end of occupational exposure (11). In each individual, a biexponential function was fitted to the data. There was a significant correlation b e t w e e n lead level in finger bone and the size of the slow component, expressed as the y-intercept. The median half-time of that c o m p a r t m e n t was 5.6 yr. However, there was a considerable interindividual range (Fig. 3). There was an uncertainty in the determination of the half-times in a particular subject. However, statistical testing s h o w e d that the variation b e t w e e n individuals was probably not the result of uncertainties, but reflected a biological interindividual variation in half-times. Lead was analyzed in vertebral bone biopsies from 27 active lead workers and 9 retired ones (9). In this total material there was a significant correlation (rs = 0.64) b e t w e e n lead levels in finger bone and vertebra, but the variation was considerable. In the active workers there was no systematic difference between levels in the two skeletal sites. However, in the exlead workers the level in vertebra was significantly lower than in finger bone. Further, there was a significant correlation between finger bone lead and blood lead in retired lead workers (rs = 0.42), but not in active ones (6). The difference in blood lead between subjects with the lowest and the highest skeletal lead concentrations was about I t~mol/L. A correlaBiological Trace Element Research
Vol. 13, 1987
BiologicalMonitoringby In Vivo XRF
245
Bone-Pb (~j/g) 1509
125-
9
O
V
o
100,
vo
ooo 9
9
ii 9
75
9 A 6
9 /
".,/;,
o 9
o
,o
9
50'
25 r
0 0
10 20 30 40 Time of employment (y)
50
Fig. 1. Relationship between time of exposure (t) and lead levels in finger bone (Bone-Pb). Closed symbols denote 75 active lead workers (circles, smelter workers; triangles pointing upward, metal moulders; triangles pointing downward, storage-battery workers; squares, demolition workers; stars, a spray painter and a worker in a paint-pigment production plant) and open ones 32 retired smelter workers. The area below the detection limit is shadowed. A function [bone - Pb = 86.8(1 - exp[ - 0.110 • t)] has been fitted to the data in active workers. tion w a s n o t e d also for vertebral b o n e lead a n d b l o o d lead (rs = 0.50; 9). H o w e v e r , in that case, a correlation w a s seen also in active w o r k e r s (rs -0.44). Biological TraceElement Research
VoL 13, 1987
Skerfving et aL
246
Bone-Pb OJg.g- 1) 200-
TRange 9
Group 2 T1/2=8.2 y
1 6
100-
6
[
T
50-
• 6
6
8 7
N
7
20-
m m
Group 1 T1/2=6.7 Y
m
,[
100
5
10
15
Time after end of exposure (y) Fig. 2. Decrease of lead levels in bone (Bone-Pb) after end of occupational exposure. Group 1 was followed for yr 0-5 after end of exposure; group 2 yr 6-13. Circles denote means of all subjects studied during that particular year. The regression lines indicate the average monoexponential decrease in each group. The half-times corresponding to these lines are given. The area below the detection limit is shadowed.
Discussion Our data show a dose-dependent (duration and intensity of exposure) accumulation of lead in bone. However, the variation in skeletal lead at a certain exposure time was considerable. This may partly be dependent upon variations in the intensity of exposure. But it should also be affected by interindividual variations in lead metabolism, e.g., turnover of the skeletal lead, which should cause differences in the accumulation pattern. It has usually been assumed that the turnover rate of skeletal lead is very slow (8,12). In accordance with this, the present data on workers temporarily removed from exposure show that there is no large, very rapid pool. However, the studies of the pattern of accumulation of lead in bone on time of exposure and the data on the decrease of lead level in
Biological Trace Element Research
VoL 13, 1987
BiologicalMonitoring by in Vivo XRF
247
B-Pb(pmol/I) 10 5.0
Subject No. 103 T1/2 4 2
~ 2.0 9
D~O~
9
9
1.0'
S~ubject No. 120 T1/2=9.5 y
0.5-
0.2"
i
i
0
i
II
i
II
1 2 3 4 5 Time after end of exposure (y)
|
6
Fig. 3. Decrease of blood lead levels (B-Pb) after end of occupational exposure in two exlead workers. A two-compartment model was fitted to the data. Biological half-times for the slow compartments are given. A "background" of 0.3 t~mol/L was subtracted from the measurement data. finger bone with time in exlead workers clearly demonstrate a rather fast turnover, the half-time being about 5-10 yr, with a remarkable consistency between both studies. This half-time is also strongly supported by the study of the elimination of blood lead after the end of exposure, which indicates a two-compartment system, the slow c o m p o n e n t of which probably corresponds to the skeleton. The data also show that the skeletal lead pool is not kinetically homogeneous. The turnover of lead in the vertebra (mainly trabecular or cancellous bone) seems to be faster than in the finger bone (mainly cortical or compact bone). However, as the decrease of blood lead, which should reflect the total skeleton, is similar to the turnover in finger-bone, the latter may be an index of the lead status of the total bone mass. The rather rapid turnover of the skeletal lead pool is also reflected by its impact on the blood lead level. A lead worker is exposed to lead not only by external absorption but, in addition, internally from the skeleton.
Biological Trace Element Research
Vol. 13, 1987
248
Skerfving et al.
In biological monitoring of lead workers by blood lead levels, this fact must be kept in mind. And this endogenous exposure continues for a long time after the end of exposure.
CADMIUM
Background Cadmium is a metal that is retained for a very long time in the kidney cortex, and may eventually cause kidney damage. During the last decades, cadmium levels in urine and blood have been used for biological monitoring of exposure. Later, in vivo neutron activation analysis (IVNAA) has been employed for measurements of cadmium levels in the total kidney (13,14). As an alternative, we have employed an XRF technique in order to minimize the absorbed radiation dose, improve the detection limit, and restrict the measurements to the toxicologically relevant part of the kidney, i.e., the cortex.
Subjects and Methods A group of 20 workers in a factory producing alkaline batteries was studied. They had an average exposure time of 16 yr (range 7-36). The air levels of cadmium in the plant at different time periods were earlier estimated on the basis of measurements, as well as on information on the production (15). Only subjects without biochemical evidence, prior to the study, of damage in the proximal part of the kidney cortex tubuli were included (~2-microglobulin in urine below 22 p,g/mmol creatinine). Cadmium levels in kidney cortex were determined by an XRF method (16), which employs plane polarized photons produced by scattering of the radiation from a therapy X-ray tube in a polymethylmethacrylate disk. The measurement time was 30 min. The mean absorbed dose to the kidney was 1.8 mGy and the total absorbed energy in the body, 0.2 mJ. The position of the kidney was determined by use of ultrasonic technique. Cadmium levels in urine were determined by a flame atomic absorption method (2) and f32-microglobulin levels in urine by radioimmune assay (Pharmacia Diagnostics AB, Uppsala, Sweden). These urinary parameters were also determined in a reference group of 20 males without occupational cadmium exposure.
Results Thirty-two determinations were performed in the 20 workers. The limit of detection was 17 I~g/g of kidney cortex (three standard deviations above background) and the precision, 23%. The average level in individual workers was 147 ~g/g (range 53-317). There was no statistically signi-
Biological Trace Element Research
Vol. 13, 1987
Biological Monitoring byln Vivo XRF
249
ficant correlation between exposure to cadmium, neither expressed as time of employment nor time-integrated exposure (l~g/m3 x yr), on the one hand and cadmium concentrations in kidney cortex on the other. When compared to a reference group, the cadmium workers had higher average levels of cadmium (5.4 vs 0.8 nmol/mmol creatine) and f~2-microglobulin (14.4 vs 6.6 ixg/mmol creatinine) in urine. In spite of the inclusion criteria, six of the workers had 132-microglobulin levels in urine above 22 lag/mmol creatinine. There were no significant correlations between levels of cadmium or [32-microglobulin in urine on the one hand and cadmium levels in the kidney on the other.
Discussion The in vivo XRF method has a detection limit sufficiently low for determination of cadmium levels in kidney cortex, even in a considerable fraction of-the population without occupational exposure (17). The levels encountered in cadmium workers were generally far higher, but similar to findings in cadmium workers by use of IVNAA (13,14). However, they were generally lower than the level of about 200 I~g/g estimated to be associated with kidney cortex damage in about one-tenth of an exposed population (18), and our data are thus in accordance with that threshold. The lack of correlation between the exposure estimates, as well as the levels of cadmium and [32-microglobulin in urine on the one hand, and cadmium levels in kidney cortex on the other, may have several explanations. In addition to different kinds of method errors, it may be caused by interindividual differences in the metabolism of cadmium (2) and/or an increase of cadmium excretion, and following decrease of kidney cadmium concentration, caused by a slight kidney damage, the presence of which was indicated by an increase of [32-microglobulin excretion in some of the subjects. Our data show that determination of the level of cadmium in kidney cortex has limitations as a means for biological monitoring of exposure. However, it could be a valuable index of risk of kidney damage; the critical level in the kidney is relatively well established, and levels in urine of cadmium and [32-microglobulin, the indices commonly in use for biological monitoring, do not mirror the level in kidney.
MERCURY
Background Different mercury compounds have a varying toxicology. From a practical point of view, the two most important species are elemental mercury vapor, which is the major risk in many occupational settings, and methyl mercury, which is present in fish. Biological Trace Element Research
VoL 13, 1987
250
Skeffving et aL
A report (19) that dentists may have elevated levels of mercury in bone and that such an increase was associated with peripheral neuropathy prompted us to perform a pilot study of bone mercury concentrations.
Subjects and Methods Five factory workers exposed to elemental mercury vapor were assayed. They had worked for 8-40 yr (mean 17) in a fluorescent-tube fitting plant or in a facility for distillation of waste amalgam from dental practices. In addition, we studied a fisherman and his wife who both consumed fish with levels of methyl mercury of about 0.5 mg/kg. Mercury in bone was determined by the same in vivo XRF technique as was used for lead (see above). Measurements were performed in finger bone as well as in the frontal bone of the skull. The detection limit was 30-50 ~g/g (three standard deviations above the background). Levels of mercury in whole blood were determined by flameless atomic absorption (20).
Results The industrial workers had an average mercury level in blood of 160 nmol/L (range 49490 nmol/L) during the period of biological monitoring of exposure (1-2 times per year for up to 11 yr). All five had mercury levels in bone below the detection limit. The two subjects exposed to methyl mercury through fish consumption both had mercury levels of 84 nmol/L in blood. None of them had detectable mercury concentrations in bone.
Discussion The industrial workers had a history of long-term exposure to elemental mercury vapor and had moderately increased blood mercury levels (1) in the range sometimes associated with slight toxic effects (4). In spite of that, the mercury concentrations in bone were below the detection limit. This is in accordance with data on the distribution of mercury in the body (21); mercury is not a bone seeker. Determination of mercury level in bone is thus not promising as a means for biological monitoring of exposure to elemental mercury vapor. Neither are bone measurements suitable for biological monitoring of methyl mercury exposure.
ACKNOWLEDGMENTS These studies were supported by grants 79-72, 82-0026, and 82-216 from the Swedish Work Environment Fund. Clinical information on some of the workers was supplied by professor Birgitta Haeger-Aronsen, Biological Trace Element Research
VoL 13, 1987
Biological Monitoring b y In Vivo X R F
251
M.D. Statistical advice was given by Mr. Jonas Ranstam, B,Sc. Excellent technical assistance was given by Mrs. Ingalill Sandell.
REFERENCES 1. S. Skerfving, Toxicology 2, 3 (1974). 2. H. Welinder, S. Skerfving, and O. Henriksen, Br. ]. Ind. Med. 34, 22 (1977). 3. H. Welinder, M. Littorin, B. Gullberg, and S. Skerfving, Scand. J. Work Environ. Hlth. 9, 397 (1983). 4. S. Skerfving, L. Ahlgren, J. -O. Christoffersson, B. Haeger-Aronsen, S. Mattsson, and A. Schhtz, Arh. Hiy,. Rada. Toksikol. 34, 277 (1976). 5. WHO, Techn. Rep. Ser. 647 (1980). 6. J. -O. Christoffersson, A. Schlitz, L. Ahlgren, B. Haeger-Aronsen, S. Mattsson, and S. Skerfving, Am. J. Ind. Med. 6, 447 (1984). 7. L. Ahlgren, K. LidOn, S. Mattsson, and S. Tejning, Scand. J. Work. Environ. Hlth. 2, 82 (1976). 8. L. Ahlgren, B. Haeger-Aronsen, S. Mattsson, A. Schutz, Br. J. Ind. Med. 37, 109 (1980). 9. A. Schlitz, J. Ranstam, and S. Skerfving, XXI International Cony, ress oll Occupational Health, Dublin, Ireland, 1984, p. 199, Abstract. 10. A. Schhtz and Skerfving, S., Scand. ]. Work Em~iron. Hlth. 3, 176 (1976). 11. A. Schlitz, S. Skerfving, B. Gullberg, and B. Haeger-Aronsen, XX International Cony, tess on Occupational Health, Cairo, Egypt, 1981, p. 144, Abstract. 12. M. B. Rabinowitz, G. W. Wetherill, and J. D. Kopple, J. Clin. Invest. 58, 260 (1977). 13. H. A. Roels, R. R. Lauwerys, J. -P. Buchet, A. Bernard, D. R. Chettle, and T. C. Harvey, Environ. Res. 26, 217 (1981). 14. K.J. Ellis, S. Yasumura, D. Vartsky, and S. H. Cohn, Fund. Appl. Tox. 3, 169 (1983). 15. E. Adamsson, Scand. J. Work Envir. Hlth. 5, 178 (1979). 16. J. -O. Christoffersson and S. Mattsson, Phys. Med. Biol. 28, 1135 (1983). 17. C. G. Elinder, T. Kjellstr6m, L. Friberg, B. Lind, and L. Linnman, Arch. Environ. Hlth. 25, 292 (1976). 18'. T. KjellstrOm, C. G. Elinder, and L. Friberg, Environ Res. 33, 284 (1984). 19. I. M. Shapiro, D. R. Cornblath, A. J. Sumner, B. Uzell, L. K. Spitz, I. I. Ship, and P. Bloch, Lancei, i, 1147 (1982). 20. A. Schtitz and S. Skerfving, Scand. J. Work Environ. Hlth. 1, 54 (1975). 21. L. Friberg and J. Vostal, Mercury in the Environment. A Toxicological and Epidemiological Appraisal, CRC, Cleveland, Ohio, 1972.
Biological Trace Element Research
Vol. 13, 1987
Biological Monitoring, by In Vivo XRF Measurements, of Occupational Exposure to Lead, Cadmium, and Mercury S T A F F A N S K E R F V 1 N G , ~'* J A N - O V E ANDREJS
CHRISTOFFERSSON, 2
SCHLITZ,' H A N S W E L I N D E R , ~ G U N N A R SP,a, NG, 3
LARS A H L G R E N , 2 A N D S O R E N M A T T S S O N 4
'Department of Occupational Medicine, University Hospital, S-221 85 Lund, Sweden; 2Department of Radiation Physics, Lund University, /rialto6 General Hospital, S-21401 ~alm6, Sweden; 3Department of Occupational Health, Kapellv~gen 2, S-5 72 O00skarshamn, Sweden; and 4Department of Radiation Physics, University of Gothenburg, S-413 45 Gothenburg, Sweden. ABSTRACT In vivo X-ray fluorescence (XRF) techniques were used for biological monitoring of lead, cadmium, and mercury. Lead accumulates in bone, the level of which may thus be used for monitoring of exposure. However, there was no close association between lead levels in bone and exposure time, partly because of differences in exposure patterns and partly, probably, because of variations in the toxicokinetics of lead. There are at least two separate bone lead compartments. The average over-all half-time is probably 5-10 yr. The finger bone level may be an index of the lead status of the total skeleton. In lead workers, the mobilization of bone lead causes an "internal" lead exposure and affects the blood lead level considerably. In cadmium workers, in vivo XRF is a sensitive and risk-free method for assessment of accumulation in kidney cortex, the critical tissue as to toxic effects; workers displayed increased levels. However, there was no clear association with duration and intensity of exposure, cadmium levels in urine, or microglobulinuria. Determinations of kidney cad*Author to whom all correspondence and reprint requests should be addressed. Biological Trace Elernent Research
241
VoL 13, 1987
Skerfving
242
et
aL
mium may add important information on the state of accumulation and, thus, risk of kidney damage. Workers exposed to elemental mercury vapor, as well as fishermen exposed to methyl mercury, had mercury levels in bone below the detection limit of the XRF method. Index Entries: Lead; bone; X-ray fluorescence; toxicokinetics, accumulation of; half-time; interindividual differences; cadmium; kidney; mercury.
INTRODUCTION The term "biological monitoring" usually denotes determination, in an index medium, of accumulated or excreted amounts of a compound. Instead of the compound itself, its metabolites may be analyzed. Alternatively, a biological effect caused by the exposure may be employed. Traditionally, the index media employed have been blood, urine, and, sometimes, hair. However, the modern methods for in vivo determination of elements in different organs by radioanalytical techniques have opened new possibilities for biological monitoring of exposure to metals. We report here some results regarding occupational exposure to lead, cadmium, and mercury. Biological monitoring may give information regarding exposure. This gives several advantages as compared to monitoring of exposure by air sampling: It is a true personal sampling, reflecting the inhaled amount of the compound. Also, it takes into consideration the fact that exposure may vary enormously at a certain air concentration, depending on the physical activity of the worker. Further, the problem of different absorption of particles of different size, often causing trouble in the evaluation of the health hazards associated with air levels, is avoided. Moreover, absorption through the gastrointestinal tract and the skin is included. Also, the biological monitoring may sometimes reflect a time-integrated exposure, thus indicating the intensity of the exposure far back in time, which may be important in evaluation of dose-effect and dose-response relationships. There are problems involved in the use of biological monitoring for evaluation of exposure. The toxicokinetics of the compound to be monitored must be known. Also, interindividual variations in kinetics may obscure the picture (1-3). In addition, biological monitoring may indicate risk of poisoning; nota bene if the rglationship between the level in the index medium and the threshold of the critical effect is known. Again, interindividual variations in metabolism and sensitivity must be known. We summarize here some data on the relationship between exposure to lead, cadmium, and mercury on the one hand, and concentrations of these metals in index media on the other, as well as studies on their toxicokinetics, including interindividual variation in metabolism Biological Trace Element Research
Vol. t3, t987
Biological~onitoring by In Vivo XRF
243
and the relationship between the metal levels in the index medium and toxic effect.
LEAD
Background Lead accumulates in the skeleton. On the basis of lead levels in different organs (4), it may be estimated that bone contains more than 90% of the body burden. Thus, concentrations of lead in bone might be useful for biological monitoring. The turnover of skeletal lead is usually assumed to be slow, and the level of lead in bone may therefore reflect the time-integrated exposure over a long time. But the evidence of a slow turnover is only circumstantial, and more information is needed. Considering the huge skeletal pool of lead, even a slow release may considerably affect the blood lead level, at present the main parameter used for biological monitoring of lead exposure (5). Also, a rapid mobilization could constitute a risk of poisoning.
Subjects and Methods In a cross-sectional study (6), 75 active lead workers (mainly smelter workers, storage battery workers, cast bronze founders, and demolition workers) were studied. In addition, we investigated 32 retired smelter workers. In a longitudinal study, 21 lead workers were followed after the end of exposure. Lead levels in finger bone were analyzed by an in vivo X-ray fluorescence (XRF) method, using 57Co as the radiation source (6-8). The detection limit is 20 ~g/g (three standard deviations above the background). The precision of the method is about 15%. In addition, lead levels were analyzed in bone biopsies obtained from the spinous process of a lumbar vertebra (9). Analysis of lead was performed by electrothermal atomic absorption after combined dry and wet ashing. The detection limit was about I ~g/g and the precision about 5%. Lead levels in blood and urine were determined by flame atomic absorption spectrometry (10). Time-integrated blood lead level was calculated in subjects who had been monitored for blood lead during their entire time (up to 11 yr) of occupational lead exposure (6). By use of the least-square method, monoexponential functions were fitted to the data on the accumulation and elimination of lead from the skeleton, and biexponential functions to the data on elimination of lead from the blood (11). Biological Trace Element Research
Vol. 13, 1987
244
Skerfving et aL
Results The median lead level in finger bone in the 75 active lead workers was 43 ~g/g; the corresponding figure in the 32 retired workers was 59 ~g/g (Fig. 1). There w a s a rise of lead levels in finger bone with increasing time of exposure. A function of the type, bone-Pb = A[1 - exp( - B x t), was fitted to the data (Fig. 1). A is a constant, B the elimination rate constant, and t the exposure time. The value of the elimination constant thus arrived at corresponds to a half-time for lead in finger bone of 6.3 yr. However, the interindividual variation was considerable; the function explained 45% of the total variance of lead levels in bone on time of exposure. In 43 subjects, the time-integrated blood lead level was compared to the lead levels in finger bone (6). There was a significant correlation, but the interindividual variation of bone lead at a certain integrated blood level was considerable. In 18 workers, w h o were temporarily removed from exposure because of high blood lead levels, determinations of lead concentration in finger bone were m a d e at the beginning as well as at the end of the exposure-free interval, which averaged 5 too. During that period, no systematic decrease of lead concentration in bone could be demonstrated. In 14 ex-lead workers, lead level in finger bone was followed after the end of occupational exposure (Fig. 2). There was a decrease in all but one. A monoexponential function was fitted to the data. The average half-time was 7 yr. In 21 ex-lead workers, blood lead level was followed for up to 13 yr after the end of occupational exposure (11). In each individual, a biexponential function was fitted to the data. There was a significant correlation b e t w e e n lead level in finger bone and the size of the slow component, expressed as the y-intercept. The median half-time of that c o m p a r t m e n t was 5.6 yr. However, there was a considerable interindividual range (Fig. 3). There was an uncertainty in the determination of the half-times in a particular subject. However, statistical testing s h o w e d that the variation b e t w e e n individuals was probably not the result of uncertainties, but reflected a biological interindividual variation in half-times. Lead was analyzed in vertebral bone biopsies from 27 active lead workers and 9 retired ones (9). In this total material there was a significant correlation (rs = 0.64) b e t w e e n lead levels in finger bone and vertebra, but the variation was considerable. In the active workers there was no systematic difference between levels in the two skeletal sites. However, in the exlead workers the level in vertebra was significantly lower than in finger bone. Further, there was a significant correlation between finger bone lead and blood lead in retired lead workers (rs = 0.42), but not in active ones (6). The difference in blood lead between subjects with the lowest and the highest skeletal lead concentrations was about I t~mol/L. A correlaBiological Trace Element Research
Vol. 13, 1987
BiologicalMonitoringby In Vivo XRF
245
Bone-Pb (~j/g) 1509
125-
9
O
V
o
100,
vo
ooo 9
9
ii 9
75
9 A 6
9 /
".,/;,
o 9
o
,o
9
50'
25 r
0 0
10 20 30 40 Time of employment (y)
50
Fig. 1. Relationship between time of exposure (t) and lead levels in finger bone (Bone-Pb). Closed symbols denote 75 active lead workers (circles, smelter workers; triangles pointing upward, metal moulders; triangles pointing downward, storage-battery workers; squares, demolition workers; stars, a spray painter and a worker in a paint-pigment production plant) and open ones 32 retired smelter workers. The area below the detection limit is shadowed. A function [bone - Pb = 86.8(1 - exp[ - 0.110 • t)] has been fitted to the data in active workers. tion w a s n o t e d also for vertebral b o n e lead a n d b l o o d lead (rs = 0.50; 9). H o w e v e r , in that case, a correlation w a s seen also in active w o r k e r s (rs -0.44). Biological TraceElement Research
VoL 13, 1987
Skerfving et aL
246
Bone-Pb OJg.g- 1) 200-
TRange 9
Group 2 T1/2=8.2 y
1 6
100-
6
[
T
50-
• 6
6
8 7
N
7
20-
m m
Group 1 T1/2=6.7 Y
m
,[
100
5
10
15
Time after end of exposure (y) Fig. 2. Decrease of lead levels in bone (Bone-Pb) after end of occupational exposure. Group 1 was followed for yr 0-5 after end of exposure; group 2 yr 6-13. Circles denote means of all subjects studied during that particular year. The regression lines indicate the average monoexponential decrease in each group. The half-times corresponding to these lines are given. The area below the detection limit is shadowed.
Discussion Our data show a dose-dependent (duration and intensity of exposure) accumulation of lead in bone. However, the variation in skeletal lead at a certain exposure time was considerable. This may partly be dependent upon variations in the intensity of exposure. But it should also be affected by interindividual variations in lead metabolism, e.g., turnover of the skeletal lead, which should cause differences in the accumulation pattern. It has usually been assumed that the turnover rate of skeletal lead is very slow (8,12). In accordance with this, the present data on workers temporarily removed from exposure show that there is no large, very rapid pool. However, the studies of the pattern of accumulation of lead in bone on time of exposure and the data on the decrease of lead level in
Biological Trace Element Research
VoL 13, 1987
BiologicalMonitoring by in Vivo XRF
247
B-Pb(pmol/I) 10 5.0
Subject No. 103 T1/2 4 2
~ 2.0 9
D~O~
9
9
1.0'
S~ubject No. 120 T1/2=9.5 y
0.5-
0.2"
i
i
0
i
II
i
II
1 2 3 4 5 Time after end of exposure (y)
|
6
Fig. 3. Decrease of blood lead levels (B-Pb) after end of occupational exposure in two exlead workers. A two-compartment model was fitted to the data. Biological half-times for the slow compartments are given. A "background" of 0.3 t~mol/L was subtracted from the measurement data. finger bone with time in exlead workers clearly demonstrate a rather fast turnover, the half-time being about 5-10 yr, with a remarkable consistency between both studies. This half-time is also strongly supported by the study of the elimination of blood lead after the end of exposure, which indicates a two-compartment system, the slow c o m p o n e n t of which probably corresponds to the skeleton. The data also show that the skeletal lead pool is not kinetically homogeneous. The turnover of lead in the vertebra (mainly trabecular or cancellous bone) seems to be faster than in the finger bone (mainly cortical or compact bone). However, as the decrease of blood lead, which should reflect the total skeleton, is similar to the turnover in finger-bone, the latter may be an index of the lead status of the total bone mass. The rather rapid turnover of the skeletal lead pool is also reflected by its impact on the blood lead level. A lead worker is exposed to lead not only by external absorption but, in addition, internally from the skeleton.
Biological Trace Element Research
Vol. 13, 1987
248
Skerfving et al.
In biological monitoring of lead workers by blood lead levels, this fact must be kept in mind. And this endogenous exposure continues for a long time after the end of exposure.
CADMIUM
Background Cadmium is a metal that is retained for a very long time in the kidney cortex, and may eventually cause kidney damage. During the last decades, cadmium levels in urine and blood have been used for biological monitoring of exposure. Later, in vivo neutron activation analysis (IVNAA) has been employed for measurements of cadmium levels in the total kidney (13,14). As an alternative, we have employed an XRF technique in order to minimize the absorbed radiation dose, improve the detection limit, and restrict the measurements to the toxicologically relevant part of the kidney, i.e., the cortex.
Subjects and Methods A group of 20 workers in a factory producing alkaline batteries was studied. They had an average exposure time of 16 yr (range 7-36). The air levels of cadmium in the plant at different time periods were earlier estimated on the basis of measurements, as well as on information on the production (15). Only subjects without biochemical evidence, prior to the study, of damage in the proximal part of the kidney cortex tubuli were included (~2-microglobulin in urine below 22 p,g/mmol creatinine). Cadmium levels in kidney cortex were determined by an XRF method (16), which employs plane polarized photons produced by scattering of the radiation from a therapy X-ray tube in a polymethylmethacrylate disk. The measurement time was 30 min. The mean absorbed dose to the kidney was 1.8 mGy and the total absorbed energy in the body, 0.2 mJ. The position of the kidney was determined by use of ultrasonic technique. Cadmium levels in urine were determined by a flame atomic absorption method (2) and f32-microglobulin levels in urine by radioimmune assay (Pharmacia Diagnostics AB, Uppsala, Sweden). These urinary parameters were also determined in a reference group of 20 males without occupational cadmium exposure.
Results Thirty-two determinations were performed in the 20 workers. The limit of detection was 17 I~g/g of kidney cortex (three standard deviations above background) and the precision, 23%. The average level in individual workers was 147 ~g/g (range 53-317). There was no statistically signi-
Biological Trace Element Research
Vol. 13, 1987
Biological Monitoring byln Vivo XRF
249
ficant correlation between exposure to cadmium, neither expressed as time of employment nor time-integrated exposure (l~g/m3 x yr), on the one hand and cadmium concentrations in kidney cortex on the other. When compared to a reference group, the cadmium workers had higher average levels of cadmium (5.4 vs 0.8 nmol/mmol creatine) and f~2-microglobulin (14.4 vs 6.6 ixg/mmol creatinine) in urine. In spite of the inclusion criteria, six of the workers had 132-microglobulin levels in urine above 22 lag/mmol creatinine. There were no significant correlations between levels of cadmium or [32-microglobulin in urine on the one hand and cadmium levels in the kidney on the other.
Discussion The in vivo XRF method has a detection limit sufficiently low for determination of cadmium levels in kidney cortex, even in a considerable fraction of-the population without occupational exposure (17). The levels encountered in cadmium workers were generally far higher, but similar to findings in cadmium workers by use of IVNAA (13,14). However, they were generally lower than the level of about 200 I~g/g estimated to be associated with kidney cortex damage in about one-tenth of an exposed population (18), and our data are thus in accordance with that threshold. The lack of correlation between the exposure estimates, as well as the levels of cadmium and [32-microglobulin in urine on the one hand, and cadmium levels in kidney cortex on the other, may have several explanations. In addition to different kinds of method errors, it may be caused by interindividual differences in the metabolism of cadmium (2) and/or an increase of cadmium excretion, and following decrease of kidney cadmium concentration, caused by a slight kidney damage, the presence of which was indicated by an increase of [32-microglobulin excretion in some of the subjects. Our data show that determination of the level of cadmium in kidney cortex has limitations as a means for biological monitoring of exposure. However, it could be a valuable index of risk of kidney damage; the critical level in the kidney is relatively well established, and levels in urine of cadmium and [32-microglobulin, the indices commonly in use for biological monitoring, do not mirror the level in kidney.
MERCURY
Background Different mercury compounds have a varying toxicology. From a practical point of view, the two most important species are elemental mercury vapor, which is the major risk in many occupational settings, and methyl mercury, which is present in fish. Biological Trace Element Research
VoL 13, 1987
250
Skeffving et aL
A report (19) that dentists may have elevated levels of mercury in bone and that such an increase was associated with peripheral neuropathy prompted us to perform a pilot study of bone mercury concentrations.
Subjects and Methods Five factory workers exposed to elemental mercury vapor were assayed. They had worked for 8-40 yr (mean 17) in a fluorescent-tube fitting plant or in a facility for distillation of waste amalgam from dental practices. In addition, we studied a fisherman and his wife who both consumed fish with levels of methyl mercury of about 0.5 mg/kg. Mercury in bone was determined by the same in vivo XRF technique as was used for lead (see above). Measurements were performed in finger bone as well as in the frontal bone of the skull. The detection limit was 30-50 ~g/g (three standard deviations above the background). Levels of mercury in whole blood were determined by flameless atomic absorption (20).
Results The industrial workers had an average mercury level in blood of 160 nmol/L (range 49490 nmol/L) during the period of biological monitoring of exposure (1-2 times per year for up to 11 yr). All five had mercury levels in bone below the detection limit. The two subjects exposed to methyl mercury through fish consumption both had mercury levels of 84 nmol/L in blood. None of them had detectable mercury concentrations in bone.
Discussion The industrial workers had a history of long-term exposure to elemental mercury vapor and had moderately increased blood mercury levels (1) in the range sometimes associated with slight toxic effects (4). In spite of that, the mercury concentrations in bone were below the detection limit. This is in accordance with data on the distribution of mercury in the body (21); mercury is not a bone seeker. Determination of mercury level in bone is thus not promising as a means for biological monitoring of exposure to elemental mercury vapor. Neither are bone measurements suitable for biological monitoring of methyl mercury exposure.
ACKNOWLEDGMENTS These studies were supported by grants 79-72, 82-0026, and 82-216 from the Swedish Work Environment Fund. Clinical information on some of the workers was supplied by professor Birgitta Haeger-Aronsen, Biological Trace Element Research
VoL 13, 1987
Biological Monitoring b y In Vivo X R F
251
M.D. Statistical advice was given by Mr. Jonas Ranstam, B,Sc. Excellent technical assistance was given by Mrs. Ingalill Sandell.
REFERENCES 1. S. Skerfving, Toxicology 2, 3 (1974). 2. H. Welinder, S. Skerfving, and O. Henriksen, Br. ]. Ind. Med. 34, 22 (1977). 3. H. Welinder, M. Littorin, B. Gullberg, and S. Skerfving, Scand. J. Work Environ. Hlth. 9, 397 (1983). 4. S. Skerfving, L. Ahlgren, J. -O. Christoffersson, B. Haeger-Aronsen, S. Mattsson, and A. Schhtz, Arh. Hiy,. Rada. Toksikol. 34, 277 (1976). 5. WHO, Techn. Rep. Ser. 647 (1980). 6. J. -O. Christoffersson, A. Schlitz, L. Ahlgren, B. Haeger-Aronsen, S. Mattsson, and S. Skerfving, Am. J. Ind. Med. 6, 447 (1984). 7. L. Ahlgren, K. LidOn, S. Mattsson, and S. Tejning, Scand. J. Work. Environ. Hlth. 2, 82 (1976). 8. L. Ahlgren, B. Haeger-Aronsen, S. Mattsson, A. Schutz, Br. J. Ind. Med. 37, 109 (1980). 9. A. Schlitz, J. Ranstam, and S. Skerfving, XXI International Cony, ress oll Occupational Health, Dublin, Ireland, 1984, p. 199, Abstract. 10. A. Schhtz and Skerfving, S., Scand. ]. Work Em~iron. Hlth. 3, 176 (1976). 11. A. Schlitz, S. Skerfving, B. Gullberg, and B. Haeger-Aronsen, XX International Cony, tess on Occupational Health, Cairo, Egypt, 1981, p. 144, Abstract. 12. M. B. Rabinowitz, G. W. Wetherill, and J. D. Kopple, J. Clin. Invest. 58, 260 (1977). 13. H. A. Roels, R. R. Lauwerys, J. -P. Buchet, A. Bernard, D. R. Chettle, and T. C. Harvey, Environ. Res. 26, 217 (1981). 14. K.J. Ellis, S. Yasumura, D. Vartsky, and S. H. Cohn, Fund. Appl. Tox. 3, 169 (1983). 15. E. Adamsson, Scand. J. Work Envir. Hlth. 5, 178 (1979). 16. J. -O. Christoffersson and S. Mattsson, Phys. Med. Biol. 28, 1135 (1983). 17. C. G. Elinder, T. Kjellstr6m, L. Friberg, B. Lind, and L. Linnman, Arch. Environ. Hlth. 25, 292 (1976). 18'. T. KjellstrOm, C. G. Elinder, and L. Friberg, Environ Res. 33, 284 (1984). 19. I. M. Shapiro, D. R. Cornblath, A. J. Sumner, B. Uzell, L. K. Spitz, I. I. Ship, and P. Bloch, Lancei, i, 1147 (1982). 20. A. Schtitz and S. Skerfving, Scand. J. Work Environ. Hlth. 1, 54 (1975). 21. L. Friberg and J. Vostal, Mercury in the Environment. A Toxicological and Epidemiological Appraisal, CRC, Cleveland, Ohio, 1972.
Biological Trace Element Research
Vol. 13, 1987
