Renal Failure

ISSN: 0886-022X (Print) 1525-6049 (Online) Journal homepage: https://www.tandfonline.com/loi/irnf20

Usefulness of Biomarkers of Exposure to Inorganic

Mercury, Lead, or Cadmium in Controlling
Occupational and Environmental Risks of
Nephrotoxicity

Harry A. Roels, Perrine Hoet & Dominique Lison

To cite this article: Harry A. Roels, Perrine Hoet & Dominique Lison (1999) Usefulness
of Biomarkers of Exposure to Inorganic Mercury, Lead, or Cadmium in Controlling
Occupational and Environmental Risks of Nephrotoxicity, Renal Failure, 21:3-4, 251-262, DOI:
10.3109/08860229909085087

To link to this article: https://doi.org/10.3109/08860229909085087

Published online: 07 Jul 2009.

Submit your article to this journal

Article views: 819

View related articles

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=irnf20
 Renal Failure, 2 1(3&4), 25 1-262 ( 1 999)

Usefulness of Biomarkers of Exposure to

Inorganic Mercury, Lead, or Cadmium in
Controlling Occupational and
Environmental Risks of Nephrotoxicity

Harry A. Roels, MSc, PBD, Perrine Hoet, MD,

MIH, and Dominique Lison, MD, MIH, PhD
Industrial Toxicology and Occupational Medicine Unit
School of Public Health
Uniuersite atholique a‘e Lauuain
Brussels, Belgium

ABSTRACT

A success-Iprevention of renal diseases induced by occupational or envi-

ronmental exposure to toxic metals such as mercury fig), lead (Pb), or cad-
mium (Cd) largely relies on the capability to detect nephrotoxic effecb at a
stage when they are still reversible or at least not yet compromising renal
function. The knowledge ofdose-effectkesponse relations has been useful to
control nephrotoxic effects of these metals through a “biologicalmonitoring
of exposure approach ”.
Chronic occupational exposure to inorganic mercury (mainly mercury va-
por) may result in renal alterations affecting both tubules and glomemli.
Most of the structural orfunctional renal changes become signijkant when
urinary mercury (HgU) exceeds 50 pg Hg/g creatinine. However, a marked
reduction ofthe urinary excretion ofprostaglandin E2 wasfound at a HgU
of35 pg Hg/g creatinine. As renal changes evidenced in moderately exposed
workers were not related to the duration o f H g exposure, it is beliwed that
those changes are reversible and mainlv the consequence of recently ab-
sorbed merculy. Thus, monitoring HgL’ is usefulf o r controlling the nephro-
toxic risk of overexposure to inorganic mercury; HgU should not exceed 50
p g Hg/g creatinine in order to prevent cytotoxic undfirnctional renal effects.
Address correspondence to: Harry A Roels, Uniti de Toxicologic Indudrielle et Mkdecine du Travail, U n i v d atholique
de Louvam, Clos Chapelle-aux-Champs 30.54,1200 Brussels, Belgium. Tel: 32 2 764 3223;Fax: 32 2 764 3328;omail
roels@oxi. ucl.ac.be.

25 1

Copyright 0 1999 by Marcel Dekker, Inc. w w w .dekker.com

252 Roels, Hoet, and Lison

Several studies on Pb workers with blood lead concentrations (PbB) usu-

ally below 70 pg Pb/dl have disclosed either no renal eflects or subclinical
changes of marginal or unknown health sign$cance. Changes in urinan, ex-
cretion of eicosanoids was not associated with deleterious consequences on
either the glomerularjltration rate (GFR) - estimatedfiom the creatinine
clearance (CCJ - or renal hernodynamics ifthe workers ’ PbB was kept below
70 pg Pb/dL. The health signifzcance of a slight renal hype$ltration state in
Pb workers is yet unknown. In terms ofPb body burden, a mean tibia Pb con-
centration of about 60 pg Pb/g bone mineral (that is 5 to 10 times the aver-
age “normal”concentration corresponding to a cumulative PbB index of 900
pg Pb/dL x year) did not affect the GFR in male workers. This conclusion
mrry not necessarilv be extrapolated to the general population, as recent
studies have disclosed inverse associations between PbB and GFR at low-
level environmental Pb exposure. 4 10-fold increase in PbB (e.g.from 3 to
40 pg Pb/dL) was associated with a reduction of 10-13 mL/min in the Ccr
and the odds ratio of having impaired renal function (viz. Cry 5th percen-
tile: 52 and 43 mumin in men and women, respectively) was 3.8 (CI
1.4-10.4; p = 0.01).However, the causal implication of Ph in this association
remains to be clarified.
The Cd concentration in urine (CdU) has been proposed as an indirect
biological indicator for Cd accumulation in the kidney. Several hiomarkers
f o r detecting nephrotoxic effects of Cd at d@rent renal sites were studied in
relation to CdC! In occupationally exposed males, the CdU thresholds ,for
significant alterations of renal markers ranged. according to the marker,
from 2.4 to 11.5 p g C d g creatinine. A threshold of I0 pg Ca7g creatinine
(corresponding to 200 p g C d g renal cortex: lhe critical Cd concentration in
the kidney) is confirmedf o r the occurrence of low-molecular-mass protein-
uria functional effect) and subsequent loss of renalfiltration reserve capac-
iq,In workers, microproteinuria wasfound reversible when reduction or ces-
sation of exposure occurred time& when tubular damage was still mild ($2-
microglobulinuria I500 pg/g creatinine) and CdU had never exceeded 20
pg Cd/g creatinine. As the predictive significance of other renal changes
(biochemical or cytotoxic) is still unknown, di seems prudent to recommend
that occupational exposure to Cd should not allow that CdU exceeds 5 pg
C d g creatinine. In groups of the general Belgian population with low-level
Cd exposure due to historicalpollution of the environment by zinc/cadmium
smelters, much lower no-adverse effect CdU levels were found, viz. 3 p g
Cd24 hours f o r low-molecular-massproteinuria and 2 pg Cd23 hours for
hypercalciuna (I 985 -1 989 CadmiBel Study).As hypercalciuria may exacer-
bate the development of osteoporosis, especially in the elderly, a CdU of 2 pg
C d g creatinine should be regarded as a measure of the maximum tolerable
internal dose of Cdf o r individuals in the general population.

Key Words: Nephrotoxicity; Mercury; Lead: Cadmium; Biomarkers of expo-

sure.
Control of Nephrotoxicity from Hg, Pb, or Cd Exposure 253

INTRODUCTION

Upon chronic exposures to inorganic forms of cadmium (Cd), lead (Pb), or mercury)&I(
nephropathy may occur which usually starts insidiously. A cascade of events may develop
leading from initial dysfunction and focal damage to a clinically detectable disease (1). ‘I’he
kidney is characterized by its ability to compensate for renal damage and for this reason
classical tests (e.g. serum creatinine or inulin clearance) are insensitive, since they only deviate
late in the cascade of damage-events when already a large part of the nephron mass is lost (2:).
It is thus of paramount importance that tools for detecting heavy metal nephrotoxicity be
sensitive enough to detect early events and ideally they should reflect subclinical reversible
changes. During the last two decades much effort has been put in the development and valida-
tion of markers of early renal effects (3,4). Such biomarkers are not diagnostic tests, but
hallmarks of early changes in the renal integrity that could later lead to clinical disease if
exposure is not reduced. Knowledge about the predictive value of these markers is particularly
useful for preventive purposes.
In order to control the nephrotoxic risk of chronic exposure to cumulative metals such as Cd,
Pb, and Hg, which are often present in industrial settings and in the general environment, im
epidemiological approach can be used to define acceptable exposure levels. This requira,
however, the knowledge of the relationship between the dose of these metals and the probability
of occurrence of adverse effects in populations at risk. Biomonitoring of the cumulative internal
dose of these metals is possible through direct in vivo measurements using neutron activation
analysis (NAA) or X-ray fluorescence (XRF) methods, such as for Cd in hdney and liver ( 5 ) ,
Pb in bone (6), and Hg in kdney (7). However, these techmques are not easlly available and
not yet well suited for routine biomonitoring of exposure in the occupational or environmental
settings. Therefore, indirect estmators are often used as surrogate biomarkexx of the cumulative
internal or target dose. viz. Cd in urine (Cdu), cumulative blood lead index (CBLI), and Hg in
urine (HgU). These surrogate biomarkers of cumulative exposure have shown satisfactmy
correlations with kidney cortex Cd in workers with no signs of Cd-inducedproteinuria (3,tibial
Pb (8), and renal Hg (7) respectively. On the other hand, currently available analyt~calm&ak
allow to determine small amounts of endogenous substances in urine and blood whch are usefd
as biomarkers of nephrotoxicity in order to assess function and integrity of speclfic nephron
segments (2,4).
Several studies conducted by our laboratory in groups of workers exposed to Cd, Pb or Hg
vapor allowed to establish dose-effect or dose-response relations which have highhghted the
usefulness of biomonitoring of exposure to these metals for controlling neplmtoxic risks m the
occupational setting. Our szudies conducted in groups of the general population environmen1:ally
exposed to low levels of Cd or Pb demonstrated, however, that h d i n g s in industrial male
workers are not necessarily applicable to the general population. The present paper gives im
overview of the most sigruficant frndings obtained by these epidemiological investigations.

EPIDEMIOLOGICALFINDINGS

Mercury

Excessive exposure to rnorganic mercury compounds either through inhalation of elemenlal

mercury vapor, ingestion of divalent mercury salts, or the use of skin-lightening cosmetics
containing inorganic mercury may entail a nephrotic syndrome (severe albuminuria) or an acute
254 Roels, Hoet, and Lison

tubular necrosis (9). As environmentalexposure to inorganic mercury is usually not a public

health issue, most of the knowledge on renal effects of inorganic mercury exposure was
obtained from studies in groups of subjects occupationally exposed to mercury vapor in
chloralkali plants. Recently, there has been much debate on the health risk of dental amalgams,
but the reports on symptoms and s i p and the few epidemiologicalstudies were inconclusive
(9). It was previously believed that renal effects of exposure to mercury vapor occued only
at doses lugherthan those associated with the onset of signs and symptoms of central nervous
system dyshction. However, numerous iuvestigations during the last two decadeshave shown
renal effects in workers with relatively low exposure levels (9). In several of these studies
signdicant associations between biomarkem of renal effects and HgU were found which allowed
to assess an HgU threshold.
In 1980, we reported on a group of 63 Hg-exposed workers (mean age 36) compared with
an age-matched control group of 88 male subjects (10). The mean duration of exposure was 7
years and HgU averaged 59 pg Hg/g matinine (range 5-200). An increasedurinay excretion
of high-molecular-mass proteins (albumin, transferrin, IgG) was indicative of an early glo-
merular dysfanction; the proximal tubular cells were also affected as shown by an increased
urinary excretion of lysosomal p-galactosidase activity. These biomarkers of tubular and
glomerular effects correlated significantly with HgU. The likelihoad of finding these nephro-
toxic effects associated with exposure to mercury vapor increased s@cantly in workers with
HgU > 50 pg Hg/g creatinine (10). Another study, carried out 5 years later in a p u p of 180
male and female workers evidenced only proximal tubular effects, viz. increases in 0-
galactosidase and urinary retinol-binding protein (RBP). Again, the prevalence of abnormal
values of these tubular markers rose sigruficantly when HgU exceeded 50 pg Hg/g creatinine
(1 1). Recently,we assessed the usefulness of a battery of about 25 renal biomarkers in the
frame of a European network study on a cohort of 44 moderately exposed chloralkali workers
(12). The mean duration of exposure was 11 years and HgU averaged 22 pg Hg/g creatinine
(range 5-90) which was less than half that m the two pxevious cohorts.Dose-response relations
showed no changes in functional biomarkers since glomerular or tubular proteinuria was not
detected. The main renal changes were indicative of tubular cytotoxicity, viz. increased leakage
of tubular antigens (BB50, BBA, HF5) and enzymes such as the brush-border lysosomal N-
acetyl-gD-glucosddase (NAG) and the S3-segment-specific intestinal alkaline phosphatase
(IAP).Sigdicant changes in biochemical indicators were also found, viz. decreases in urinary
eicosanoids (pGE2, PGFZa,TXB2), glycosaminoglycans, and kallikrern. Most of these renal
effects were found in workas excreting more than 50 pg Hg/g creatiuine corroboratmg thus the
previous proposal of an adequate HgU threshold (10,ll). However, a sigdicant reduction in
the urinary excretion of eicosanoids, especially PGE2, was noted at HgU values as low as 35
pg Hg/g creatinine.None of the changes in this less exposed group were related to duration of
exposure, which supports the view that renal changes mducedby moderate exposure to mercury
vapor are reversible and mainly the consequence of receatly absorbed elemental mercury
It may be concluded from these three studies that except a few biochemical changes such as
a decrease m urinary pH and a reduction m PGEz excretion of whch the health sigruficance is
not yet well understood, the nephrotoxic response to moderate occupational exposure to
mercury vapor does not involve cytotoxic and functional alterationswhen HgU remainsbelow
50 pg Hg/g creatinine. Them is, however, an increased probability of cytotoxic effects at the
proximal tubuli (e.g. enzymmia and i n c m in urinary tubular antigens) and functionalchanges
(e.g. g l o m d a r and tubular proteinuria, increase in serum p2-microglobulin)when mercury
vapor exposure chronically entails HgU levels above 50 pg Hg/g creatinine. Disrupted renal
handling of low and high-molecular-massproteins did not lead to clirucally si@cant altera-
Control of Nephrotoxicity from Hg, Pb, or Cd Exposure 2:55

tions of the kidney function, as tubular and/or glomerular proteinuria were shown to Ix
reversible 9 to 12 months after cessation of exposure (1 1,13). The reversibility time course:
seems to parallel the elimination lunetics of renal m e r c q . In a chloraykali worker wil3 an HgU
of 200 pg Hg/g creatinine, for example, the concentration of mercury in urine wdl slowly
decrease to reach “acceptable” values (10-20 pg Hg/g creatinine) in about one year after
removal fiom mercury vapor exposure, because the elimination half-tme of HgU is on average:
about 90 days ( 14).

Lead

Chronic massive exposure to lead may cause progressive tubuloinierstitial nephropathy thal:
develops insidiously and often leads to kidney failure ( 15- 18). The level of Pb exposure whchi
may be associated with early adverse renal effects is however uncertain because incipient lead1
nephropathy is Micult to detect due to the lack of an appropriate bloodborne or urinary
biomarker reflecting an early effect on the renal interstitial tissue. In chronically exposed Pbh
workers, a decline of glomerular filtration rate (GFR) has been reported in cases with long
standing exposure and lead in blood (PbB) exceeding 70 pg Pb/dL (16,17). A reduced GFR
however, indicates that Pb nephropathy has already reached an irreversible stage. Hence, to
better d e h e the occupational exposure level without early adverse renal effects, a large number
of cross-sectional studies were conducted during the last two decades in lead workers whose:
duration of exposure was less than 30 years and whose PbB usually remained below 60 iq;
Pb/dL (for a review, ref 19). The majority of these studies and also the European network study,
involving the large renal biomarker battery. disclosed a lack of association between the urinary
excretion of renal markers of tubular or glomerular toxicity and biomarkers of Pb exposure, viz
PbB, zinc-protoporphyrin in blood (ZPP), or tibia lead (610.19-22). Some studies hwe:
revealed simicant changes in the urinary excretion of the distal tubular enzyme kallikre:ini
(23,24), NAG (25-28). or eicosanoids, viz. two vasodilators 6-keto-PGF1, and PGE2 which
decreased and a vasoconstrictor thromboxane B2 which increased (2;!). but the c h c a l signifi-
cance of these changes is not yet clear.
The crucial question is: What is the predictive value of these renal changes and should they
be taken into account to assess the no-adverse-effect level of lead on the kidney?
First, the specifcity of the reduced urinary activity of kallikrein (23,24) and the increased1
excretion of NAG in urine (25-28) found in these studies can be questioned, because conconli-
tant albeit slight exposure to Cd, which frequently occurs in lead industries, has not been1
considered as a possible confounder. Indeed, contrast analysis in groups of Cd and Pb workers
showed CdU but not PbB as determinant of urinary kalllkrein activity (29), and as to urinary
NAG activity. its association with biomarkers of lead exposure (viz. PbB, ZPP, or tibia lead)
was abolished after adjustment for CdU in stepwise multiple regression analysis (6,30). The
changes in urinary excretion of eicosanoids might, however, reflect a disturbance of their
synthesis in the kidney (22). As the renal production of eicosanoids may play a role in the
regulation of the GFR, we advanced the hypothesis of a possible effect of Pb on renal h e m a -
namics before the perturbation of other renal biomarkers. Therefore. in order to assess whether
these eicosanoid changes had any impact on the renal hemodynamic response to an acute oraI
protein load, we compared 76 male workers (mean age 44, range 29-56) moderately exposed
to lead with 68 male controls matched for age and socioeconomic state (6). The lead workers
were exposed on average for 18 years (range 6-36), their hstorical PbBs rarely exceeded ’70
pg Pb/dL, and they showed a threefold higher body burden of lead than the controls as estimated
by in vzw XRF measurements of lead in tibia (66 vs. 2 1 pg Pb/g bone mineral). The baselint:
256 Roels, Hoet, and Lison

and peak C a averaged 116 and 132 mL/min in the control p u p against 121 and 141 mL/min
in the Pb-exposed group, respectively.All the subjects had a normal basehe CC,. In stepwise
multqle regression analysis, both baselme and peak Ca were inversely associated with age (p
< 0.01), whereas a modest but positive association emerged only with tibial bone lead ( j<~
0.05). These findings suggest that the cumulative Pb exposure is associated with a slight state
of renal hyperliltration. In conclusion. it seems thus unlikely that the underlying changes in
urinary eicosanoids have deleterious consequences on renal hemodynamics in this group of
moderately exposed Pb workers. Th~sstudy suggested that a mean tibia lead concentration of
about 60 pg Pb/g bone mineral (that is about 5 to 10 times the average “normal” concentration)
is not Uely to affect the GFR and supports the view that adverse renal changes are unlikely to
occur in adult male workers when the individual PbB is kept below 70 pg Pb/dL during the
whole occopational career (3 1).
Previous studies from the USA ( 15) and Australia ( 18) have shown that environmental Pb
exposure may severely compromise renal function and entail important implications for public
health. As to the possibility of renal impairment with low-level lead exposure in humans, there
is growing evidence (32,33) suggestingthat the no-adverse-effectPbB level of 70 pg Pb/dL for
occupational lead exposures cannot be extrapolated to the general population whtch is usually
exposed to lower lead levels. A random sample of about 2000 adults (20 pg Pbig bone 88 years
old) from the Belgian population was investigated (32) using the database generated in the
frame of the C a m e l Study (see also below “Cadmium”). After adjustment for covariates
(age, body-mass index, and diuretic treatment) si&icant inverse relations were found in men
and women between the measured Ccr and PbB or ZPP. A tenfold increase in PbB (e.g. from
4 to 40 pg Pb/dL) was found associated with a reduction in Ccr of 10 and 13 mL/min in men
and women, respectively. In men and women without diabetes and analgesic or diuretic drug
use, the 5th pementile of the Ccrwas 52 and 43 &min, respectively. The subjects with lower
Ccr were classified as having impaired renal function (men 48/965, women 6211016). In
multiple logistic-regression analysis with adjustments for covariates (viz. age, the presence of
diabetes, and the use of analgesic or diuretic drugs). the probability of having an impaired renal
function was positively associated with PbB and in the total study population the odds ratio for
a tenfold increase in PbB was 3.8 (CI 1.4-10.4:~= 0.01). Since several studies suggested that
renal insufficiency itself does not raise bone lead (34) or PbB (35,36), our results (32) and those
of Payton et al. (33) lend support to the hypothesis that low-level environmental exposure to
lead may reduce the GFR. However, the critical cumulative Pb dose in the general population,
e.g. bone lead concentration or cumulative blood lead index, which is associated with the onset
of a marginal impairment of the GFR remains to be assessed (37).

Cadmium

Cadmium is a very cumulative toxic metal that under conditions of chronic occupational or
environmental exposure has the kidney as its critical target organ in the human organism.
Epidermological studies performed in the 1970s and early 1980s on groups of male Cd workers
have shown that Cd may interfere with the renal handling of plasma-derived proteins. T ~ wasB
usually characterized by microproteinuria due to impairment of the tubular reabsorption of low-
molecular-mass proteins, e.g. prmicroglobulin @&) and RBP (38,39). An isolated glomeru-
lar effect with increased permeability of high-molecular-mass proteins, e.g. albumin and
transfenin, was less commonly found (39,40). In male Cd workers, the risk of microprotemuria
increased sqpficantly when CdU regularly exceeded 10 pg Cdg crealinine which corresponds
to a critical Cd concentration in renal cortex of about 200 pg Cd/g wet weight as estimated by
in vivo neutron activation analysis (5.38.41).
Control of Nephrotoxicity from Hg, Pb, or Cd Exposure ;!5 7

In the frame of the collaborative network-project, involving laboratories of 5 European

countries. we applied recently the battery of 25 renal tests on a cohort of 37 male Cd workas
(CdU: 2 to 16 pg Cdg creatinine) and an age-matched control group (n = 43; CdU < 2 pg Cdjg
creatinine). The aim was to assess the usefulness of various urinary and bloodborne analytes
as potential biomarkers of Cd toxicity on distinct nephron segments (42). The battery comprised
functional markers (e.g.creatinine and p2-M in serum; low and high-molecular-massproteinis
in urine, and urinary calcium), cytoroxicity markers (e.g. tubular antigens or enzymes in urine),
and biochemical markers (e.g . glycosaminoglycans, kallikein. sialic acids, and eicosanoids in
urine). Cd exposure was found to cause a broad spectrum of site-specific effects on the kidney
with sigruficant alterations in various markers of nephrotoxicity. Dose-responserelations imd
logistic regressions between the renal markers and CdU showed that the probability of abnor-
mal values sigdkantly increased with CdU in a dose-dependent manner. Three main groups
of CdU thresholds could be identified on the basis of various urincuy biomarkers: one around
2 pg C d g creatinine associated with biochemical alterations (6-keto-PGF1, and sialic acids),
a second around 4 pg Cd/g creatinine associated with cytotoxic effects (renal brush-border
antigen BBA, IAP.and NAG) and glomerular barrier dysfunction (albumin and transferrin), ,anid
a third around 10 pg C d g creatinine associated with dysfunction of the tubular reabsorphn
(microproteinuria: (p2-M and RBP), hypercalciuria, and changes in other biomarkers (rmd
brush-border antigen HF5 and glycosaminoglycans).
A critical question is whether or nor there enough mformation as to the prdctive value of
these renal biomarkers to propose a meaningful biological exposure limit value for CdU.Earlier
retrospective findmgs showed that severe microproteinuria was irreversible in retired Cd
workers and associated with raising serum creatinine (43) that in some workers turned into mil-
stage renal insuflkiency (unpublished data). We have carried out three studies to better asrieijs
the health si&icance of Cd-induced microproteinuria in male Cd workers.
The first study was a 5-year prospective study conducted in 23 Cd workers removed froim
exposure because of the discovery of microproteinuria (44). They 'were exposed to Cd for i!5
years on average and at the time of the first examination the mean age of the group was 59
(46-68 years), the mean ( S E N CdU amounted to 22.2 (2.9) pg CdL, the geometric mean of
urinary 02-M and RBP amounted to 1770 and 1570 pgiL, respectively, and serum creatirluie
was normal (< 1.4 mg/dL) in all subjects. except two (2.0 and 2.2 mg/dL). At the end of tlie
study. urinary 02-Mand RBP were increased by about 50 and 30S6. respectively. Moreover,
serum creatinine si&icantly increased with time indicating a progressive reduction of tlie
GFR. which was estimated to decrease five times more rapidly than what could be expected due
to aging alone. Elevated microproteinuria predicts thus an exacerbation of the age-related
decline of the GFR and. hence. it should be regarded as an early adverse health effect.
The previous finding raised the question whether a CdU threshold value of 10 pg Cd/g
creatinine prevents not only the occurrence of microproteinuria. but also the loss of nephon
mass. In other words, does an increased CdU not yet sufficient to m+ the urinary excretion
of plasma-derived proteins impair the renal filtration reserve capacity? In a second study vve
have estimated the GFR in Cd workers without (n = 3 1) or with (n = 12) increased micropro-
teinuria. viz. urinary B2-M and RBP > 300 pglg creatinine (45). The subjects in both grolulps
aged 50 to 64 (mean 55 years). had normal serum creatinine (< 1.4 mg/dL,), and the geometric
mean (range) of CdU was 4.7 (2.1-8.8) and 11.1 (5.8-21.7) pg Cd/g cEatinine, respeetive1,y.
GFR was estimated as the Ccrunder baseline conditions and after an acute oral protem load to
assess the hyperfiltration capacity of the kidney. The baseline Ccr 'was normal in both grctu!ps
(mean 116 d m i n ) . The Ccr after protein load, however, failed to rise in the group with
25 8 Roels, Hoet, and Lison

microproteinuria (mean 114 d m i n ) and remained sqyuficantly lower than that in the group
without microproteinuria (mean 124 mL/min). The mean peak Ccr of the latter group was
s d a r to that of an age-matched control group (n = 35; CdU < 2 pg Cd/g creatinine). In
conclusion, thts study showed that the filtration r e m e capacity of the kidney is lost when
elevated microprotemuria is present, but there was no functional impairment at a renal Cd
burden that had not yet caused microproteinuria. This study vddated thus the CdU estimate of
10 pg W g creahine as biological exposure l imitto prevent the o c c m c e of micrqmteiuuria
in male Cd workers.
The third study aimed at a more precise evaluation of the time trend of Cd-induced micro-
proteinuria by assessing its evolution in 32 Cd workers according to criteria of CdU (cutoff 10
pg Cdg creatinine) and severity of the microproteinuria at the time that the exposure was
substantially reduced or had ceased (46).The finding that 15 workers (63%) with CdU > 10 pg
Cdg creatinine showed a urine PTM exceeding the upper reference h u t of 300 pg/g creatinine
corroborated our earlier finding that the risk of abnormal microproteinuria dramatically
increased when CdU regularly exceeded 10 pg Cd/ g creatinine. When reduction of Cd
exposure took place and urine P2-M did not exceed 300 pg/g cxeatinine, the risk of developing
tubular dysfundion at a later stage was low even in cases with fustorical CdU values occasion-
ally > 10 but always < 20 pg Cd/g creatinine. There is indication that the tubulotoxic effect of
Cd is reversible, provided that the historical CdU values had never exceeded 20 pg Cd/g
creatinine and Urinary PTM was mild (< 1,500 pg/g creatinjne) at the time the Cd exp- was
reduced. When severe urinary P2-M (> 1,500 pg/g creatinine) was found in combinationwith
historical CdU values exceeding 20 pg Cd/g creatinine, Cd-induced tubular dysfunction was
progressive in spite of reduction or cessation of Cd exposure.
These three studies on the health signuficance of microproteinuriaconfirm thus the estimate
of 10 pg C d g creatinine as the biological limit value of CdU above which the probability is
high that Cd-induced microproteinuria develops. Hence, tubular proteinuria should be consid-
ered as an adverse effect of Cd exposure, because it can lead to irreversible renal damage
associated with an exacerbation of the age-related decline in GFR and a decxease in the filtration
reserve capacity. Since the health si@icance of renal changes other than microproteinuria is
stdl unknown and no treatment is presently available to remove Cd from its storage sites, the
adoption by the American Conference of Governmental Industrial Hygiemsts (ACGIH) of 5 pg
Cdg cxeatinine as biological expo- index (BEI) for Cd seems logical (47). However, studies
on the predictive value of renal changes other than microproteinmia are urgently needed to
assess the validity of this BEI.
As to environmental Cd exposure in the general population, a more stringent guideline is
justified as fust suggested by earlier pilot-studies in Belgium (48-50) and as conclusively
shown by the 1985-1989 CadmBel Study on the possible health effects of low-level Cd
exposure in the Belgmn population at large (5 1-55). The vast cross-sectionalepidemiological
investigation on the renal function c o m m a sample of about 1700 adults (males and females
aged 20-80 years) who had never been occupationally exposed to Cd, F'b, or Hg, and who were
randomly recruited in two less polluted areas (maland urban) and two other areas (rural and
urban) characterizedby environmentalpollution due to the activities of various zinc/cadmium
smelters in the past ( 5 1,52).After allowing for major covariates, e.g. age and smoking habits,
the 24 h urinary Cd excretion averaged 25% higher in women than in men and was found in
both genders 20 to 60 YOhigher in the more polluted rural and urban areas (53). In stepwise
multiple regression analysis, only the 24 h excretions of urinary markers of renal tubulotoxicity
(viz. calcium, NAG, RBP, prM, and amino acids) were si&cantly and positively associated
with CdU (51). In the rural areas (about 1100 subjects), a doubling of the Cd excretion was
Control of Nephrotoxicity from Hg, Pb, or Cd Exposure 25 9

Table 1

Biological exposure thresholdsfor controlling health-signijkant nephrotoxic effects in adults

chronicalIy exposed to either moderate occupational or low-level environmental levels of inorganic:
mercury, lead, or cadmium.

Occupational exposure Environmental exposure

(healthy males; age 20-60 ) (general population)
Hg in urine Not relevant
50 pg Hg/g crea!inine
reversible eflect:
glomerular and/or tubular
proteinuria

Lead (Pb) Pb in blood Pb in blood

70 pg PbldL (31) ?
decrease in GFR ? #decreasein GFR ?
no validated ear& biomarker no validated early biornarker

Cadmium (Cd) Cd in urine Cd in urine

10 pg C d g creatinine 2 pg C d g creatinine
critical adverse eflect: critical adverse eflect:
microprote inuria hypercalciuria

5 pg C d g creatinine
BEl adopted tyACGIH (47)

GFR: g l o m d a r fitration rate: BEI: Enological exposure index, ACGIH: American Conference of
Governmental Iudustrial Hygienists.

associated with a 4, 7, and 9% increase in the excretion of RBP, urine P2-M, and NAG
respectively, and living 1 km closer to the nearest smelter with a 1 to 2% increase in their
excretion against 3% for the Cd excretion (54). After adjustment for covariates. 1ogisti;c
regression models between CdU and each of these five urinary biomarkers of tubular effects
have shown that the likelihood of abnormal biomarker values (subclinical increases in excre-
tion) were to occur with a probability of 10% at a urinary Cd excretion of 1.9,2.7.2.9,3.1,and
4.3 pg Cd24 h for calcium. NAG, RBP, urine P2-M. and amino acid-j, respectively (52). These
results suggested that the no-adverse-effect level of urinary Cd in the general populatiatn
approximates 3 pg Cdg creatinine for microproteinuria (RBP and urine (j2-M),which is ibt
variance with 10 pg Cdg creatinine for adult male Cd workers. This stdimg difference might
illustrate the impact of selection biases operating in industrial worker cohorts, such as gender
(usually males), restricted age gmup (no elderly), self-selection bias, healthy worker effect, and
pre-employment medical examination preventing potentially more susceptible subjects to be
hired for jobs involving exposure to cadmium, lead, or other nephrotoxic agents. A 5-year
follow-up of the most heavily exposed subgroup in the C a m e l Study (rural area) indicated
however that the mild changes in microproteinuria were not predictive of an increased risk of
renal dysfunction (manuscript in preparation).
260 Roeis, Hoet, and Lison

The C a m e l Study revealed also a dose (CdU)-dependent calciuria and suggested an

increased pmalence of hypercalciuria when CdU exceeded 2 pg Cd/g creatinine, especially in
post-menopausal women (52,55). This hypercalciuria should be considered as an early adverse
tubulotoxic effect, because it may exacerbate the development of osteoporosis,especially in the
elderly (56). We have therefore proposed that a CdU value of 2 pg Cd/g creatinine should be
regarded as a measure of the maximum tolerable internal dose of Cd for individuals from the
general population. It may be concluded that in the early 1990s. the Cd excretion in Belgium
exceeded this threshold value in about 10%of the general population and reached 20% in the
rural area historically polluted by Cd from nonferrous smelters. In thts area. a clear-cut impact
of a preventive action to decrease the Cd transfer from the environment to the inhabitants was
observed in a follow-up study about 5 years later, because the Cd concentration had decreased
by about 30% in blood and 15% in urine: the decrease was less in subjects living closer to the
smelters and in pre-menopausal women (57).

CONCLUSION

The results of the various epidemiologicalinvestigations performed since the 1970s by our
laboratory indicate that in Belgium the efforts made by industries and the health authorities to
reduce occupational and environmental exposure to inorganic mercury, lead. and cadmium are
fully j u d i e d . Table 1 summarizesthe guidelines for a practical and sound biological monitor-
ing of exposure to these heavy metals in orda to control adverse nephrotoxic effects at an early
stage and ensure an adequate health prevention in the occupational setting and the general
population.

REFERENCES

1. Rom WN: Environmental and OccupahonalMedicme. 2nd ed. Boston, LittleBrown and Company, 1992. pp.
735-772.
2. Price RG, Taylor SA, Chivers I, et al: Development and validation ofnew screeningtestsfor nephrdoxic &&.
HurnanExp Toxicol 15 (Suppl 1): SlO-S19. 1996.
3. Lauwesys RR.Banard A: Early detection ofnepbrotoxicef€e& of industrial chemicals: state ofthe att and &we
prospecis. Am JIndMed 11:275-285,1987.
4. Mueller PW, Price RG, Finn WF: New approaches for detecting thresholds of human nephrotoxicitv using
cadmium as an example. Environ Health Perspect 106:227-230. 1998.
5. Roels HA Lauwerys RR.Buchct JF', et al: In MVO measurement of liver and kidney cadmium in workers exposed
to this metal : Its significance with respect to cadmium in blood and urine. Environ Res 26:2 17-240.198 1.
6. Roels H, Lauwerys R, K d g s J, et al: Renal b d i o n and hyperfiltration capacity in lead smelter workers with
high bone lead. OccupEnvironMed 51505-512,1994.
7. B6qesson J, Barregard L, SBllsten G, et al: In wvo XRF analysis of merauy: the relation between ooncatrations
inthekidney andtheurine. PhysMedBiol40:413-426,1995.
8. Roels H, Konings J, Green S, et al: Time-integrated blood lead concertration is a valid surrogate for estimating
the cumulativelead dose assgsed by tibia1 lead measurement. Environ Res 69:75-82. 1995.
9. WHO-IPCS: Environmental Health Criteria 118, Inorganic Mercury,.Geneva, WHO. 1991. pp 84-107.
10. Buchet JF', R d s H, b a r d A Lauwerys R Assessment ofrenal fundion ofworkem exposed to inorganic lead.
cadmium ormercuryvapor.JOccupMed22:741-750, 1980.
11. Roels H, Gemart JP,Lauwerys R et al: Surveillance of workers exposed to mercury vapour: validation of a
previously proposed biological threshold limit value for merauy concgdrationin urine. Am J IndMed 7:45-71,
1985.
12. Cardenas A Roels H, Bernard AM, et al: Markem of early renal changes induced by industrial pollutants. I.
Application to workers exposed to mercury vapour. B r J IndMed 50:17-27, 1993.
Control of Nephrotoxicity from Hg, Pb, or Cd Exposure 26 1

13. Stewad W K Guirgis HA. Sandason J, Taylor W: Urinary mercury excreti'm and proteinuria in pathology
laboratory staff Br J I n d M e d 34:26-31, 1977.
14. Roels HA, Boeckx M, Ceulemans E, Lauwerys RR: Urinary excretion of merimy after occupational exposure
to macury vapour and influence ofthe chdatmgagmt meso-2,3-dimercaptosua5nic acid (DMSA). Br JIndlMcd
48:247-253. 1991.
15. C d a JC: Clinical manitbtations andtherapy of acute lead intoxication due to the inggtion of illicitly distillid
alcohol. Ann Internhfed 59:707-715. 1963.
16. Lilis R Gavrilescxl N, Nestorescu B, et al: Nephropathy in chronic lead poisoning. Br J IndMed 25:196-202,
1968.
17. Wedeem RP. Maesaka JK,Weiner B, et al: Occupational leadnephropath?;.Am J M e d 59: 630-641, 19751.
18. Inglls JA, Henderson D A Emerson BT: The pathology and pathogenesis of chronic lead nephropathy occurbig
in Qumsland. JPafhOl 124:65-76. 1978.
19. k a r t JP,Bemard A, Lauwerys R: Assessment ofthyroid,testes, k i h q and aidcnomic nervous system funC%iim
in lead-exposed workers. Intilrch Uccup Environ Health 64:49-57, 1992.
20. Omae I(, Sakurai 9 Higashi T, et al: No adverse effects of lead on renal fundion in lead-exposed workers. Ind
Health 23:77-83. 1990.
21. Gehardsscm L, Chettle DR. Englysi V. et al: Kidney &&s m long term exposed lead smelter workers. Br J Ind
hfed 49:186-192, 1992.
22. Cardenas A Roels H, Bernard AM, d al: Markers of early renal changes induced by industrial pollutants. IN.
Application to workers exposed to lead. Br J IndMed 50:28-36, 1993.
23. Boscolo P, Porielli G. C&&i G, et al: Urinary kallikrein adivity of workers exposed to lead. Br J IndA4ed
35:226-229. 1978.
24. ChmielniLka J, Ko&a-Smtmka E, Szymanska JA: Arginase and kallikrein ai;tivities as b i d a n i c a l indias of
occupational exposureto lead. BrJZndMed 38:175-178, 1981.
25. Meyer BR, Fischbein A, Rosenman K, el al: Increased urinary enzyme excretion in workers exposed to
nephrotoxic chemicals. Am J S f e d 76:989-998. 1984.
26. Vmchoor hL Wibowo A Herber F, et al: Influence of ooxpational low-level lead exposure on renal parametas.
Am JIndMed 12:341-351. 1987.
27. Coratelli P, Giannattasio M. Lomonte C, et al: Enzymuria to deted tubular injury in workers exposed to lrad :
a 12 month follow-up. Contrih ,\'ephroZ 68:207-211, 1988.
28. Chia E;s. MWi 4 Tan C, ei al: Urinary N-acetyl-6-D-glucosaminidaseactivity in workers exposed to inorganic
lead. Occup EnvironMed 51:125-129, 1994.
29. Roels H.4, L a u w q s RR. Buchet JP. et al: LJrinary kallikrein adivity in workers exposed to cadmium, lead, or
mercury vapour. Br J Indhfed 47:331-337, 1990.
30. Roels H. Bernard A Lauways R: Urinary N-acBy1-8-D-glumsaminidase (NAG) and exposure to inorganic leid
(Reply). Occup Environ ,Wed 52:286-287. 1995.
31. Council Diredive of 28 July 1982 on the proteclion of workers fiom the risks related to exposure to metallic l a d
and its ionic compounds at work (first individual Diredive within the meanmg of Article 8 of Diredive
8011107:EEC) (821605.EEC). UficialJ E u r Community 25: No L247' 12-21 (23.8.82), 1982.
32. Staessen J A Lauwerys RR. Bud& JF'. et al: Impairment of renal fundion withi increasingblood lead cmcatra-
tions in thegeneralpopulation. .VEngZJhfed 327:151-156, 1992.
33. Paytm hL Hu H, Sparrow D, Weis ST: Low-level lead exposure and renal function in the nonnative aging study.
A m JEpidemiol 140:821-829, 1994.
34. Emmason BT. Lecky DS: The lead content of bone in subjects without recognised past lead exposure and in
patients with renal disease. AustralnsAnnhfed 12:139-142, 1963.
35. Campbell BC. Elliott HL, Meredith PA: Lead exposure and renal failure: does renal insufficiency influence lead
kinetics? ToxicolLett 9:121-124. 1981.
36. Batuman V, Landy E, Masaka JK,W e d m RP: Contribution of leadto hypatension with renal impairment. N
Engl Jl\fed 309: 17-21.1983.
37. American Council on SLience and Health (ACSH):LeadandHuman Health. New Yo* ACSY 1997,pp 1-37.
38. WHO-IPCS: Environmental Health Criteria 134, Cadmium. Geneva. WHO, 1992, pp 132-146.
39. Bernard A Buchrt JP,Roels H. et al: Renal excretion of proteins and enzymes m workers exposed to cadmium.
Eiir J Clin Invest 9: 11-22. 1979.
40. Bernard AM, Roels H, Cardmas A, Lauwerys R: Assesanent of urinary protein 1 a n d t r d m i n as early maikrrs
of cadmiumnephrotoxicit?;.Br J I n d M e d 47:559-565, 1990.
41. Roels H. Lauwerys R, D a r d w e AN: The critical level of cadmium in human rmal codex: a re-evaluaticn.
Toxic01 Len 15:357-360. 1983.
42. Roels H. Bernard AM, Cardenas A, et al: Markers of early renal changes induced by industrial pollutants. 111.
Application to workers exposed to cadmium. Br J IndMed 50: 37-48. 1993.
262 Roels, Hoet, and Lison

43. Rods H, Dpbpng J, BlactKt JP,d al: Evolution ofcadmiummducedrenal dysfundMl in workas moved fian
exposure. ScandJ WorkEmron Health 8:191-200,1982.
44. Roels HA, h e r y s RR,Buchd JP,t i al: Health sigifimce of cadmium induced renal dy&mim~:a fieyear
follow up. Br J IndMed 46:755-764,1989.
45. Rods HA, Lauweays RR,Banard AM, et al: kxsnxnt ofthe filtration m e capacity ofthe kihey in workas
exposedtocadmium. BrJIndMed48: 365-374,1991.
46. Roels HA, Van U e FJ, O v e r s t e p s M, d al: Reversibilay of m i q r & i n u r i a in cadmium workers with
incipienttubular dysfimdion after reduction of exposure. Am J Ind Med 3 1 : 645452,1997.
47. Amaican Confixence of Govemmatal Industrial Hygienists (ACGIH): Supplement: Cadmlum and Inorganic
Compounds. Biological Exposure Indices,Cinchmati, ACGIH, 1996.
48. Roels HA, Lauwerys R, Bu&d JP, W a r d A Environmental exposureto cadnium and renal iimdon of aged
women m three areas of Belgium. Environ Res 241 17-130,1981,
49. Lauweays R, M a t s P:hvironmmtal pollution by cadmium and mortality fiom renal diseases.Lancer i: 383,
1981.
50. Lauwerys R, Hardy R, Job M, et al: Envirmmtal pollution by cadmium and cadmium body burden:an autopsy
study. Toxrcol Left 23:287-289, 1984.
51. Lauwerys R, Amery A, Bernard A, et al: Health effieds of envirmmtal exposweto cadmium: ObjeAves, design
and organization ofthe Cadmibel study: a uosssertonal morbidity study camed out in Belgium &om 1985 to
1989. Enwron Health Perspect 87: 283-289,1990,
52. Wl&d JP,Lauwerys R, Rods H, d al: Renal d e d s of cadmium body burden ofthe grnesal population Lancet
336:699-702, 1990.
53. SMor FA, Rondia DJ, Claeys FD, et a1 Impad ofenvirmental cadmium pollution on cadmium exposure and
body burden. Arch Environ Health 47:347-353, 1992.
54. Sta- JA, Lauwerys RR, Ide G, d al: Renal fimdion and historical-ne I cadmium pollution h m
zinc smelten. Lancet 343:1523-1527, 1994.
55. Sta- J, Amery A, h a r d A, et al: Effeds of exposureto cadmium on calcium metabolism : a population
study.BrJIndMed45:710-714,1991.
56. Sta- JA, Roels HA, Emelianov D, et al: EnVimrrmartal exposureto cadmium,forearm bone density and risk
offiactures: a prospeUivepopulation study in Belgium. Lancet 353:1140-1144,1999.
57. Sta- JA, Kuznekova T,Emelianov D, et al: hviramental exposure to cadmium and conventional and
ambulatoryblood pressures in a prospedive population study.Am J Hypertens (submitted).