Journal of Exposure Science and Environmental Epidemiology (2011) 21, 247–261
r 2011 Nature America, Inc. All rights reserved 1559-0631/11
www.nature.com/jes
Application of biological monitoring for exposure assessment following
chemical incidents: A procedure for decision making
PAUL T.J. SCHEEPERSa, PETER M.J. BOSb, JOKE KONINGSa, NICOLE A.H. JANSSENb AND LINDA GRIEVINKb
a
Department of Epidemiology, Biostatistics and HTA, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
National Institute for Public Health and the Environment, Bilthoven, The Netherlands
b
Determination of the level of exposure during and after a chemical incident is crucial for the assessment of public health risks and for appropriate medical
treatment, as well as for subsequent health studies that may be part of disaster management. Immediately after such an incident, there is usually no
opportunity to collect reliable quantitative information on personal exposures and environmental concentrations may fall below detectable levels shortly
after the incident has passed. However, many substances persist longer in biological tissues and thus biological monitoring strategies may have the
potential to support exposure assessment, as part of health studies, even after the acute phase of a chemical incident is over. Reported successful
applications involve very persistent chemical substances such as protein adducts and include those rare cases in which biological tissues were collected
within a few hours after an incident. The persistence of a biomarker in biological tissues, the mechanism of toxicity, and the sensitivity of the analysis of
a biomarker were identified as the key parameters to support a decision on the feasibility and usefulness of biological monitoring to be applied after an
incident involving the release of hazardous chemicals. These input parameters could be retrieved from published methods on applications of biomarkers.
Methods for rapid decision making on the usefulness and feasibility of using biological monitoring are needed. In this contribution, a stepwise procedure
for taking such a decision is proposed. The persistence of a biomarker in biological tissues, the mechanism of toxicity, and the sensitivity of the analysis of
a biomarker were identified as the key parameters to support such a decision. The procedure proposed for decision making is illustrated by case studies
based on two documented chemical incidents in the Netherlands.
Journal of Exposure Science and Environmental Epidemiology (2011) 21, 247–261; doi:10.1038/jes.2010.4; published online 24 March 2010
Keywords: analytical methods, biological monitoring, environmental exposures, inhalation exposure, chemical incidents.
Introduction
An acute chemical incident, or disaster, may be defined as ‘‘a
situation in which people are potentially exposed to hazards
to which they are vulnerable, with resulting public concern,
and the possibility of immediate or delayed risks to health’’
(Wisner and Adams, 2002). Such events include fires,
explosions, leakages, and accidental or deliberate releases of
toxic substances that may cause illness, injury, disability, or
death. Determination of the level of exposure during, and
after, such incidents is crucial for the assessment of public
health risks and appropriate medical treatment, as well as
for subsequent health studies (WHO, 1997, 2009). Health
studies that include exposure assessment might contribute to
effective disaster management (Bongers et al., 2008; Van den
Berg et al., 2008).
1. Address all correspondence to: Dr. Paul T.J. Scheepers, Research Lab
Molecular Epidemiology, Department Epidemiology, Biostatistics and
HTA, Postal code 133, Radboud University Nijmegen Medical Centre,
PO Box 9101, 6500 HB Nijmegen, The Netherlands.
Tel.: þ 31 24 3616878. Fax: þ 31 24 3613505.
E-mail: [email protected]
Received 9 September 2009; accepted 8 January 2010; published online 24
March 2010
Immediately after a chemical incident, exposure assessment is often limited to a restricted number of environmental
measurements, usually conducted by the fire department.
These measurements can be supplemented later by measurements conducted by experts such as hazmat teams. The
(often short) duration of the primary exposure, the limited
time available to deal with exposure issues (relative to other
priorities such as the saving of lives), and the poor
information on the identity of the toxic chemicals in an early
phase of an incident are particular problems that can result in
discrepancies between the desired and attainable exposure
data (WHO, 1997). The quality of information collected in
this early phase after an incident may be barely sufficient to
distinguish between areas of high and low exposure. Hence,
uncertainties remain with regard to the actual levels to which
individual subjects have been exposed. Many substances
persist longer in biological tissues for which reason biological
monitoring provides a good alternative for characterizing
personal internal exposure to chemical substances, integrating uptake by different routes of exposure.
Within the present context, biological monitoring is
defined as the standardized and repeated systematic
collection, pretreatment, storage, and analysis of body tissues
to assess the internal dose of a xenobiotic substance by
analysis of the parent substance and/or a product of
Biomonitoring following chemical incidents
Scheepers et al.
biotransformation. An appealing element in the use of
products of biotransformation as biomarkers is that the
bioavailability and metabolic bio-activation of chemicals is
taken into account, including inter-individual variability in
genetic constitution and acquired traits.
In those cases, in particular, in which it is not possible to
collect environmental samples during an incident, for
example, because of administrative or technical problems,
the analysis of biological tissues may be the only remaining
alternative to support exposure assessment. The use of
biomarkers to estimate exposure should therefore always be
considered in health studies after disasters (WHO, 1997,
2009). Examples of specific aims of applying biological
monitoring as a tool in exposure assessment are presented in
Table 1.
Decisions on the implementation of biological monitoring
can be placed within the broader framework of health studies
after disasters. The main objectives of disaster health
outcome assessment include providing information (a) at
the individual level, (b) at the group level of the affected
population, as well as (c) at the societal level. From the
perspective of the individual victim, the goal of the
information collected could be to optimize the medical
treatment of that individual. At the group level, the outcome
of biological monitoring may contribute to public health. On
a societal level, health studies may also serve as a signal of
recognition of the problems of survivors. For them, it may be
important to know that their possible exposure to toxic
substances is being taken seriously. If enhanced exposure can
be ruled out, it might have a reassuring effect and may
prevent speculations about possible exposure to toxic
substances. Moreover, it could be seen as an intervention
in itself (Van den Berg et al., 2008).
A decision to apply biological monitoring needs to be
taken as soon as possible after an incident by the emergency
Table 1. Justification for analysis of body tissues after chemical
incident.
Aim
Confirmation of internal exposure to one
or more substances
Relate an exposure to observed clinical
symptoms
Support medical care, for example,
supportive treatment or use of antidote
Establishing the cause of death in a
post-mortem analysis
Assessment of body burden as an
exposure estimate related to persisting
health complaints (to support
injury/liability claims)
Perform a health risk assessment
a
Qualitative
Quantitative
x
xa
x
x
x
x
x
x
x
For those substance with a background in the general population.
248
response authorities. A method to support such rapid
decision making is needed (Bongers et al., 2008). The aim
of this study is to define a stepwise procedure to predict the
usefulness and feasibility of applying biological monitoring
after a chemical incident. We will identify the parameters
needed for making a timely decision and provide this
information available for 15 relevant chemical substances.
The use of the procedure will be elaborated in two case
studies extrapolated from documented incidents in the
Netherlands.
Methods
Basic Considerations
Within the present context, a basic point of departure is that
biological monitoring may be useful in determining whether
any significant exposure has occurred, especially if the
predicted exposure is expected to be sufficiently high to
induce adverse health effects. As will be elaborated below,
this can be assessed by use of intervention values for
emergency response (IVERs). The most relevant IVERs
are set for different exposure durations (generally from
10 min to 8 h). Therefore, after a chemical incident a
biomarker is useful if it can be used as approximation of
the (time-weighted average) ambient exposure during the
chemical incident and be related to possible (adverse) health
effects through the use of these IVERs. Hitherto, the
following input for evaluation and decision making for the
usefulness and feasibility of biological monitoring in a given
exposure situation can be defined:
1. Toxico-kinetic parameters, such as the biological half-life
of the biomarker and the pattern of excretion, are
necessary for a quantitative retrospective exposure assessment.
2. Toxicity mechanism-based criteria: retrospective assessment of the environmental exposure will only be relevant
if the exposure is expected to have been high enough to
induce significant adverse health effects.
3. Analytical criteria: an adequate analytical method with
sufficient accuracy and precision is needed for a quantitative determination of the biomarker level.
These criteria will be further discussed below.
Toxico-Kinetic Parameters
The possibility of detecting a chemical substance at a certain
specific time point after cessation of the exposure is
determined by the persistence of the biomarker in biological
tissues. Depending on the type of biomarker, the pattern of
elimination may vary; for example, from blood many parent
substances and their metabolites will follow a first-order
pattern of elimination (a logarithmic pattern of initial rapid
decay, followed by a tail-shaped slow decrease). A similar
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
pattern is expected in urine but this is attenuated by retention
of urine in the bladder. For biomarkers that remain
intracellular, such as hemoglobin adducts, the elimination
will be much slower and follow a linear decay with a half-life
dependent on the lifespan of the cell population (zero-order
kinetics).
Toxicity Mechanism-Based Criteria
Several programs have been initiated to facilitate rapid
decision making by emergency response units in case of
chemical incidents. In these programs IVERs are used as
action values. Within the present context, the terminology of
the acute exposure guideline levels (AEGLs) program (NRC,
2001; Krewski et al., 2004) will be used as these are
internationally most widely used. To establish these IVERs
the toxicity profile of a chemical is evaluated and three air
concentrations (i.e., AEGL-1, AEGL-2, and AEGL-3,
representing action levels) are derived that mark the
transitions between four categories of increasing health effect
(see Supplementary Information, Table 1).
Below the AEGL-1, no relevant health effects are to be
expected and the public would not or barely notice the
presence of a chemical contaminant in the air (detectability).
Between AEGL-1 and AEGL-2, some slight effects may
occur, such as slight eye irritation, but no further severe
irreversible health effects are anticipated (discomfort).
Between AEGL-2 and AEGL-3, exposure might result in
severe and/or irreversible health effects, or in effects that
could impair escape reactions (e.g., narcosis and severe eye
irritation) or (disability). Exposures above AEGL-3 are life
threatening.
AEGL-2 is the most important action level and because of
its definition (i.e., transition level for health-threatening
exposure) it can be used as a key parameter for the
determination of the usefulness of biological monitoring.
This is straight forward when the AEGL-2 is based on a
systemic health effect because biomarker levels also relate to
systemic availability of a toxic substance and/or intermediate.
The AEGL-2 can also be based on local effects such as upper
tract irritation but even then it might still be possible that
systemic effects for which biological monitoring is useful
occur at higher exposure concentrations between the AEGL2 and the AEGL-3.
Analytical Criteria
If a validated analytical method is available and the optimal
sampling strategy is determined, it is useful to verify whether
the expected biomarker values are at detectable levels at the
time of sample collection. For this, the predicted biomarker
concentration can be compared with the limit of detection
(LOD) or, for quantification purposes, the predicted value
can be compared with the limit of quantification (LOQ). If
no information on the LOQ is provided, three times the value
of the LOD was used for practical reasons.
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Scheepers et al.
Selection of Chemical Substances that may be Involved
in Chemical Incidents
The relevance of chemical substances for the Netherlands was
assessed based on the Register Risk Situations Hazardous
Substances (Dutch: Register Risicosituaties Gevaarlijke
Stoffen, RRGS) that contains the frequency of occurrence
of chemical substances as reported by the municipalities (this
is a legal requirement by 2008). This registry is made public
in the format of risk maps (www.risicokaart.nl). Some
substances were added because they may be important
constituents of process emissions and fires, such as acrolein,
hydrogen cyanide, hexavalent chromium in chromates, and
polycyclic aromatic hydrocarbons (PAH) in airborne and
particulate emissions. Four of the substances (acrylonitrile,
ethylene oxide, hydrogen fluoride, and hydrogen sulfide)
were also selected as acute exposure threshold level (AETL)
case study substances (Trainor et al. 2006).
Obtaining Key Parameters for 15 Selected Chemicals
Relevant chemical properties and absorption, distribution,
metabolism, and elimination (ADME) parameters were
mostly obtained from bibliographical databases (Pubmed,
Medline, Excerpta Medica). Some Internet sources (RTECS,
TOXNET, GESTIS, IARC Monographs, ATSDR, and
the Health Council of the Netherlands, the ACGIH BEI
documentation and DFG MAK Kommission documentation) were consulted. The pattern of elimination is ruled by
the half-life of the biomarker that reflects the persistence of
the biomarker in body tissue. If for a biomarker more than
one half-life was reported, we selected the longest reported
so-called terminal half-life. For some biomarkers no human
data could be retrieved from literature. If no human data
were available, animal data were used as an approximation.
The elimination half-life should therefore be used as an
indicative value, rather than a value representative for the
population. Most data were based on adults. Only for dioxin,
data were also available for children.
The sensitivity of the chemical analysis of the biomarker
can be valued from the reported LOD or LOQ. In this
overview, the lowest value for these parameters was reported.
This was normally also the value that was reported most
recently. These data were supplemented with a declaration of
LOD/LOQ values by laboratories that provide routine
analysis of biomarkers.
Incident Scenarios
Incident scenarios that were used are based on documented
incidents in the Netherlands. The benzene emission (case 1)
was related to a spill that occurred in the Rotterdam port on
Tuesday, 18 March 2003 at 0815 hours. In reality, the
atmosphere was slightly unstable with temperatures of 4 1C
at 10 m and 2.2 1C at 200 m (stability class B2), with a
ground temperature of 7 1C. For the purpose of this study a
fictitious worst case was assumed with a much higher outdoor
249
Scheepers et al.
air and ground temperature of 25 1C. This caused the
emission of benzene vapor to increase from 130 kg/min to
309 kg/min, causing a threefold higher release (from 7789 kg
to 18,498 kg) after 1 h.
The acrylonitrile incident (case 2) occurred in Amersfoort
on Tuesday, 20 August 2002, when at 1103 hours a spill was
discovered. The tank wagon had been leaking since
approximately 0500 hours. In reality, only a small unknown
quantity leaked from the tank load. In the extrapolated
fictitious scenario we just assumed that 50,000 of the total
load of 56,000 kg of the load had been spilled, causing an
emission of 9.5 kg/min. This was calculated using Areal
Locations of Hazardous Atmospheres (ALOHA) software
version 5.4.1.2 (EPA, 2009b).
Results
Stepwise Procedure for Decision Making
On the basis of the abovementioned considerations a stepwise
procedure is proposed to assess the usefulness and feasibility
of biological monitoring after a chemical incident (Figure 1).
The first step is based on the premise that in case of an
accidental chemical release, biological monitoring will be
useful if it can be anticipated that there is a reasonable
certainty that the AEGL-2 has been exceeded, because only
in such a case significant adverse health effects may be
expected. However, there may be occasions in which it is not
possible to predict whether the AEGL-2 has been exceeded
or occasions in which there is significant public concern about
the exposure. In these cases, biological monitoring may still
be considered if steps 3–6 are fulfilled (see below).
The second step addresses the situation that the AEGL-2
can also be based on local effects such as eye or upper
respiratory tract irritation. Internal exposure levels are
usually not related to these kinds of local effects and
estimation of the internal exposure by biological monitoring
is not useful in such cases. However, in these cases one should
verify whether systemic effects can be expected once the
AEGL-2 is exceeded (step 2a). As a rule of thumb, if
the exposure is estimated to be at least two to three times the
AEGL-2 (step 2b), it can be anticipated that these
systemic effects may occur and one should continue with
step 3.
In step 3 the most suitable biomarker should be selected,
based on specificity to the substance of exposure. For 15
priority substances we have searched the literature for
different options for biomarkers (Table 2). For most
biomarkers the half-lives are known (step 4). If the half-life
of the biomarker is not determined empirically, it may be
derived using quantitative structure activity relationships
(QSARs). The chemical substances that were selected to
study the feasibility of using biological monitoring in the case
of a chemical incident are listed in Table 2. This selection
250
Biomonitoring following chemical incidents
contains 12 singular substances and 3 substances that are
part of a group or mixture. Furthermore, the selection
contains 10 organic substances and 5 are inorganic
substances. Of the inorganic substances, three are metals.
The substances represent a great variety of health end points.
Important local end points are local irritancy and pulmonary
edema and systemic health end points include neurological
effects, vascular shock, hemolysis, effects on the gastrointestinal tract and cardio-dysrythmia, effects on metabolism,
reproduction toxicity, mutagenicity, and carcinogenicity. The
list includes seven substances that are classified as known
human carcinogens (IARC category 1). For two chemicals
(dioxin and PAH) no information was available in the
registry because they are not usual industrial or commercial
products.
An adequate analytical method with sufficient accuracy
and precision is needed for a quantitative estimate (step 5). In
step 6 it is determined how long after the end of exposure the
biomarker can still be detected. The estimation of the critical
period of time between the end of the exposure period and
the time of sample collection ts is based on a calculated
biomarker level higher than the LOQ of the analytical
method (see Appendix A) or higher than the background
biomarker level in the general population (if this is
substantially higher than the LOQ).
It is useful to determine which value of ts is the minimum
time that should be available for the preparation of
proper sample collection. In the Netherlands, the minimum
value of ts is estimated to be 72 h for a population-based
study. The time needed to collect biological materials
from a subgroup such as first responders may be much
shorter, especially if previous arrangements have been made,
such as availability of the sample collection materials. It is
noted here that in the Netherlands there are no ethical
requirements to apply biological monitoring as long as the
result of biological monitoring is in line with providing the
best health care for the victims. As outlined in the
introduction, reliable information on the identity and
magnitude of exposure can, and will, provide information
that is in the interest of an adequate treatment and public
health care.
In Table 3, the input parameters discussed above are
presented for the selection of 15 priority chemicals in the
Netherlands.
Using the Stepwise Procedure
The stepwise approach will be illustrated by two cases that
are realistic extrapolations from documented chemical
incidents in the Netherlands. These cases were selected to
include two biomarkers that follow either zero-order, or firstorder, kinetics in their pattern of elimination. This provides
an opportunity to show differences in the equations and
calculations needed to predict the time window for collection
of biological tissues.
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
Scheepers et al.
Figure 1. Stepwise procedure for decision making on the application of biological monitoring after a chemical incident (see text for further details).
Fictitious Case 1: Benzene Spill Rotterdam Harbor
Early in the morning a chemical spill was reported in the
Rotterdam Botlek harbor area. As a result of a technical
failure, 400 m3 of crude benzene (approximately 90%
benzene and 10% toluene) was spilled in an open-air
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
reservoir of approximately 3,500 m2. The pool was
10–15 cm deep. The fire department covered the surface of
the spill with foam to decrease dispersion of organic vapors
into the air. The wind direction was north with a speed of
2 m/s. A first measurement was carried out by the regional
251
Biomonitoring following chemical incidents
Scheepers et al.
Table 2. Selection of 15 substances for which the feasibility of applying biological monitoring after a chemical incident was determined.
Substance
Acrolein
Acrylonitrile
Arsenic and arsine
Benzene and gasoline
Cadmium
Chromium trioxide +
chromic VI acida
Dioxine, TCDD
Ethylene oxide
Hydrogen + potassium
cyanide
Hydrogen fluoride
Methyl bromide
Polycyclic aromatic
hydrocarbons
Styrene
Toluene
Xylene + m-xylene +
o-xylene
CAS
107028
107131
7440382
71432+86290815
7440439
AGW
Effects at AGW
(mg/m3)a (if the substance has no AGW the toxicity is described in general terms)
0.5
50
1
500
1,000
F
1333820+7738945
F
1746016
F
75218
100
74908+151508
10
7664393
74839
130498292
20
200
F
100425
108883
1330207/108383/
95476
1,000
1,000
1,000
IARC3; irritant
IARC2B; irritant; reprotoxic
IARC1; abdominal pain, vomiting, and diarrhea, vascular shock;
hemolytic (arsine)
IARC1; neurotoxic
IARC1; acute chemical pneumonitis and edema; nausea vomiting and
abdominal pain
IARC1; strong irritation and ulceration of skin and respiratory tract;
irritation of GI tract, vomiting diarrhea, acute kidney damage
IARC1; chloracne, immunological effects; multiple effects due to
enzyme induction
IARC1; strong irritation of skin and respiratory tract, pulmonary edema;
neutotoxic; reprotoxic
IARC–; inhibition of oxygen metabolism resulting in CNS and heart failure
IARC–; irritant
IARC3; irritant
IARC1-2A-2B-3 (benzoa[a]pyrene has been classified as a human
carcinogen in a recent re-evaluation); phototoxic, mutagenic
IARC2B; irritant; neurotoxic
IARC3; neurotoxic, muscle weakness
IARC3; irritant; neurotoxic
Number
of
locationsb
3
23
3
514
F
13
F
11
7
15
3
F
16
60
23
IARC–, no classification by International Agency on Research for Cancer available; IARC1, carcinogenic to humans; IARC2A, probably carcinogenic to
humans; IARC2B, possibly carcinogenic to humans; IARC3, not classifiable as to it carcinogenicity in humans; F Not reported.
a
Dutch IVER similar to the AEGL-2, see Table 1 of Supplemental Material.
b
Number of locations in which the chemical substance is stored or used in the Netherlands (based on the RRGS registry of 2008).
environmental office using a portable photo ionization
detector (PID, MiniRae, Rae Systems, San Jose, CA,
USA). During the first and second day, concentrations of
around 300 p.p.m. were measured at 200 m downwind from
the incident location. For this case, it is assumed that on the
first day eight workers and four policemen were exposed for
B8 h without wearing effective respiratory protection.
The indicative measurement of 300 p.p.m. indicates that
the interim AEGL-2 of benzene for an 8-h exposure period
of 200 p.p.m. is exceeded (step 1: yes). The AEGL-2 is based
on depression of the central nervous system, which is a
relevant end point (step 2; EPA, 2009a). For this case,
S-phenylmercapturic acid (SPMA) in urine is the biomarker
of preference because of its sensitivity and specificity to
benzene exposure (step 3, see Supplementary Information,
Table 2). The predicted level of SPMA at the end of an 8-h
exposure period is extrapolated using a regression equation
describing the relationship between inhalation exposure and
excreted levels of SPMA in workers (see Appendix A). Using
this equation, an SPMA excretion of E9000 mg/l was
calculated to correspond to the AEGL-2 of 200 p.p.m.; a
documented fast elimination corresponded to a half-life of
9.0±4.5 h (step 4). The LOQ for SPMA in urine is 1 mg/l but
252
the 95 percentile in the general population is higher: 7.3 mg/l
(Scheepers, 2009) (step 5). On the basis of these data it is
calculated that spot urine samples should be collected within
3–4 days after the end of exposure (step 6, see Appendix A).
Collection of multiple spot samples is recommended if a
quantitative exposure estimate is required (step 7). For
interpretation, there are sufficient human data on SPMA to
take into account possible confounders such as background
exposures due to active smoking and air pollution. Besides
supporting exposure assessment, the outcome of biological
monitoring may support treatment of health effects in
individuals such as on CNS and blood cells and support
decisions related to the possible treatment of these health
effects.
Fictitious Case 2: Acrylonitrile Release from a Tank
Wagon in Amersfoort
Early in the morning, health authorities were notified of
several health complaints of railway personnel and inhabitants of the residential area in Amersfoort. The complaints
were of a ‘‘strong, irritant, onion smell’’ and 21 persons were
admitted to hospital, of whom 12 patients were admitted to
the intensive care unit in a critical condition. It took some
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
Scheepers et al.
Table 3. Key properties of the biomarkers
Substance
Biomarkers
Half-life (h)
Reference
Measuring
principle
Acrolein
3-Hydroxypropylmercapturic acid
N-2-Cyanoethylene valine
adducts in blood
2-Cyanoethylmercapturic
acid (CEMA) in urine
Thiocyanate in urine
Total As in urine
414 daysa
Carmella et al. (2007)
LC-MS/MS
74 daysb
Bader and Wrbitzky
GC-MS
(2006)
Jakubowski et al. (1987) GC-FID
Acrylonitrile
Arsenic and arsine
Benzene and
gasoline
Cadmium
Hydrogen cyanide
and potassium
cyanide
Hydrogen fluoride
Methyl bromide
F
0.9
Carmella et al. (2007)
2 pmol/g
globin
F
F
Bergmark et al. (1997)
1
Schettgen et al. (2009)
2.7±1.1 daysb Schulz et al. (1979)
38 daysb
Pomroy et al. (1980)
GC-MS
ICP-MS
F
0.006
1
0.02
Shibata et al. (2004)
Heitland and Köster
(2008)
Heitland and Köster
(2008)
Scheepers (2009)
38 daysb
Pomroy et al. (1980)
ICP-MS
0.03
0.1
15b
GC-FID
0.1
1
Benzene in whole blood
8
Nomiyama and
Nomiyama (1974)
Baselt (2000)
GC-MS
0.5
1.0
S-phenylmercapturic acid
45±4b
GC-MS
F
1
Angerer and Schaller
(1985)
Hoppe (2009)e
Trans, trans-muconic acid
5.1±2.3b
HPLC-UV
F
25
Hoppe (2009)e
Cadmium in urine
10–20 years
ICP-MS
0.01
0.03
GF-AAS
F
0.1
ICP-MS
ICP-MS
0.03
0.02
0.09
0.06
GF-AAS
F
0.3
GF-AAS
F
0.6
GF-AAS
F
0.2
HRGCHRMS
1 pg/g
lipid
F
F
Haufroid et al. (2007)
HRGCHRMS
LC-MS
Walker et al. (1992)
GC-MS
1.8 pmol/g
globin
Baud et al. (1996)
GC-NPD
GC-MS
F
20
0.7
50
Shibata et al. (2004)
Odoul (1994)
Schulz et al. (1979)
(see serum)
Pierre et al. (1995)
Ekstrand and
Ehrnebo (1983)
(see plasma)
GC-ECD
HPLC-UV
GC-MS
ISEFIA
50
1,000
500
3.0
F
Shiono et al. (1991)
Shibata et al. (2004)
Kage et al. (2008)
Itai et al. (2001)
Boogaard and
van Sittert (1995)
Boogaard and
van Sittert (1995)
Lauwerys and
Hoet (2001)
7.4–16.0 yearsb Jarup et al. (1983)
730 daysb,c
Schaller et al. (2007)
Chromium in erythrocytes
126 daysb,c
Chromium in plasma
3–5 yearsb,c
TCDD in whole blood
0.4 (0.36–0.43) Leung et al. (2006)
years (children)
7.8 (5.8–9.7)
Geyer et al. (2002)
years (adults)b
B0.5–1 year
Schecter et al. (1998)
TCDD in breast milk
Ethylene oxide
LOQ Reference
(mg/l)
Inorganic As III and
As V in urine
Benzene in exhaled air
Cadmium in whole blood
Chromium trioxide+ Chromium in urine
chromic VI acida
Dioxine, TCDD
7–9
LOD
(mg/l)
Hydroxyethyl-mercapturic
o5b
acid in urine
N-2-Hydroxyethylene-valine o126 days
globin adduct in whole
blood
Cyanide in whole blood
1.14 (95% CI
0.841.80)
Cyanide in serum
Thiocyanate in urine
Fluoride in urine
Fluoride in plasma
2.7±1.1 daysb
(see serum)
15–18b
5.78b
Fluoride in whole blood
(see plasma)
Bromide in urine
10.5–14 daysb
Bromide in serum
Bromide in plasma
10–12 daysb
10–12 daysb
Not confirmed,
Lewalter et al. (1985)
Petersen et al. (2000)
Ryan and Baumann
(1999)
Sangster et al. (1983)
Rauws (1983);
Sangster et al. (1983)
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
GC-MS
F
Heitland and Köster
(2006)
Heitland and Köster
(2006)
McKelvey et al. (2007)
Heitland and Köster
(2006)
Heitland and Köster
(2006)
Heitland and Köster
(2006)
Heitland and Köster
(2006)
Stanley et al. (1986)
1 pg/g Schecter et al. (1998)
lipid
0.5 Schettgen et al. (2006)
Ahn and Shin (2006)
F
F
500
Kage et al. (2008)
GC-ECD
10
F
F
ICP-MS
ICP-MS
ICP-MS
20
5
52
F
F
F
Kawai et al. (1997)
Michigami et al. (1989)
Quinones et al. (2006)
Allain et al. (1990)
253
Biomonitoring following chemical incidents
Scheepers et al.
Table 3. Continued
Substance
Polycyclic aromatic
hydrocarbons (PAH)
Styrene
Toluene
Xylene
Biomarkers
Half-life (h)
Reference
Measuring
principle
Bromide in whole blood
12 days
ATSDR (1992)
ICP-MS
S-methylcystein albumin
adducts in whole blood
S-methylcystein Hb adducts
in whole blood
1-Hydroxypyrene in urine
20 days
Müler et al. (1994)
126 days
Müler et al. (1994)
16 daysb
Jongeneelen et al.
(1988)
HPLCPC-Flu
HPLCPC-Flu
HPLC-Flu
Styrene in whole blood
13.0
ATSDR (2007)
N-(1-Hydroxy-2-phenyl-ethyl)
valine Hb adducts in whole
blood
Mandelic acid in urine
Phenyl glyoxylic acid (PGA)
in urine
Toluene in exhaled air
Toluene in whole blood
LOD
(mg/l)
1.5
5
200 fmol/g
F
Heitland and Köster
(2008)
Müler et al. (1994)
200 fmol/g
globin
2
F
Müler et al. (1994)
F
Jongeneelen et al. (1987)
GC-MS
SPMEGC-MS
GC-MS
0.1
0.030
0.2
F
Hoppe (2009)d
Chambers et al. (2006)
0.5
1.0
1 pmol/g
globin
Angerer and Schaller
(1985)
Teixeira et al. (2008)
126 days
Rueff et al. (2009)
GC-MS
17–25
11
Symanski et al. (2001)
Symanski et al. (2001)
GC-FID
GC-FID
F
F
18.7b
21.2b
Brugnone et al. (1983)
Brugnone et al. (1983)
GC-FID
GC-MS
0.1
0.5
Hippuric acid in urine
7.3±3.8b
Hasegawa et al. (1983)
Ortho-cresol in urine
7.4±2.3b
Hasegawa et al. (1983)
HPLC-UV
GC-FID
GC-MS
Xylene in exhaled air
Xylene in whole blood
20b
20b
ACGIH (2001a)
ACGIH (2001a)
Methylhippuric
acid in urine
30.1 (range
16.1–48.4)b
Engström and
Riihimäki (1988)
LOQ Reference
(mg/l)
GC-FID
SPME
GC-MS
GC-MS
HPLCUV
2,000
F
5
0.1
50
10,000 Kezić et al. (2000)
1,000 Kezić et al. (2000)
1
1.0
Scheepers (2009)
Angerer and Schaller
(1985)
F Kataoka et al. (1991)
1,000 Weber (1992)
Angerer and Kramer
(1997)
1
Scheepers (2009)
F Chambers et al. (2006)
0.5
1.0
10,000
F
Angerer and Schaller
(1985)
Kataoka et al. (1991)
a
At 4 weeks after cessation of smoking, 78% of initial level was still detected in smokers.
Longest reported terminal half-life of excretion as determined in humans.
Determined in workers inhaling welding fumes.
d
Hoppe (2009), personal communication.
b
c
time before the probable cause of the smell was found.
According to railway authorities these serious health
complaints were possibly attributable to a spill of acrylonitrile from a railway wagon on the central railway station in
Amersfoort. The sirens were immediately set off to alert the
public. Messages were broadcasted on the emergency
channels on radio and TV to advise the public in the
residential area to stay indoors and keep windows and doors
closed. The warning reached the public apparently too late to
prevent further serious injuries, because approximately 1 h
after the start of the incident, the number of persons admitted
to hospital with serious complaints of eye irritation and upper
airway complaints reached 60. At 0945 hours, the fire
department covered the spill area with foam to reduce
evaporation. It is estimated that during the morning
approximately 50,000 kg of the product leaked from a
railway wagon on the switchyard of the railway station.
254
The product formed a circular puddle stretching approximately 9 m in each direction around the leaking tank wagon.
The vapor cloud was carried to the residential area by a weak
southwestern wind (1 m/s). The toxic vapor cloud reached
the first residents at 175–200 m.
Owing to the late warning, a 1-h exposure to acrylonitrile is
considered realistic because inhabitants were not evacuated
and were not able to close windows and shut off mechanical
ventilation at an early stage of the incident. Furthermore, it is
estimated that the proposed AEGL-2 value for a 1-h exposure
of 57 p.p.m. was exceeded in a significant part of the
residential area (step 1). This is based on the assumption that
the total mass of 50 tonnes of the product would cause a
puddle of 250 m2 and the ground temperature would be 12 1C.
Using ALOHA it was calculated that the total release would
amount to 547 kg, resulting in an AEGL-2 contour extending
132 m from the source. For this case, it is proposed to consider
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
Scheepers et al.
Table 4. Key factors in the success of an adequate response in terms of application of biological monitoring as a tool in exposure assessment.
Aspect
Success factors
Complicating factors
Biomarker
Availability of reliable human data on kinetics
of the biomarker
Availability of biological materials at an early
stage after the incident
The onset and duration of the exposure are
being recorded for each individual
The biomarker has an intrinsic chemical stability
No human data; sometimes animal data can be used as an
approximation
Sometimes samples are collected at hospitals but method of collection
and storage may not be adequate
No exact data available on the onset and/or duration of exposures
Sample collection
Exposure
Persistence of
biomarker
Ethics issues
Funding
Study can be within an existing ethically approved
framework
Funding prearranged
biological monitoring of acrylonitrile exposure for inhabitants
of the residential area who were present within the AEGL-2
contour. The AEGL-2 is based on a local effect (step 2A) but
the AEGL-3 is 100 p.p.m., which is within a factor 2 of the
AEGL-2 (EPA, 2007). Step 2B also applies for the residents
but within a smaller contour (step 2B). The preferred
biomarker is the acrylonitrile hemoglobin (N-2-cyanoethylvaline) adduct in blood because of its persistence, sensitivity, and
specificity for exposure to acrylonitrile (step 3, see Supplementary Information, Table 3). Chemically stable adducts
normally follow zero-order kinetics with a half-life of 0.5 times
the lifespan of erythrocytes (126 days in humans) (step 4).
Human data were used to extrapolate from the AEGL-2 to a
predicted adduct level at the end of the exposure incident of
E8142 mg/l blood at the end of the incident (see Appendix A).
The LOQ of this biomarker is 0.05 mg/l blood (step 5). Using
zero-order kinetics it can be calculated that blood samples
could be collected over a period of up to 150 days (step 6, see
Appendix A). It is further recommended to conduct repeated
blood sampling (e.g., once every month for 3–5 months) to be
able to reconstruct exposures retrospectively in individuals
with a relatively high level of adducts in the first blood sample
(step 7). This allows correction for increased adduct levels in
smokers and possible inter-individual differences in toxicokinetic parameters. Biomarker results may also support
decisions for medical treatment of those victims.
Discussion
The experiences with applying biological monitoring methods
in previously reported chemical incidents will be discussed to
determine the potential usefulness of biological monitoring.
The strengths and weaknesses of using a formal procedure
for decision making will also be discussed. The use of the
input parameters is a main issue for discussion. Issues related
to the implementation of biological monitoring strategies
(sampling collection strategy and interpretation of results)
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
There is an unknown loss of the biomarker substance
Time-consuming procedure for ethics approval
Not clear which authority should support the study
will be indicated briefly and some guidelines formulated as a
preparatory tool for dealing with chemical incidents. Last but
not least, the implication of biological monitoring for risk
management and risk communication will be discussed.
Success Factors
Biological monitoring of mercury, methylmercury, PCBs,
and dioxin after chemical incidents proved successful because
of the persistence of these substances in body tissues
(Bertazzi, 1991; Ackermann-Liebrich et al., 1992; Aubin
et al., 1994; Bertazzi et al., 1998; Needham et al., 1999;
Clewell et al., 2008; see Supplementary Information, Table 4
for a literature overview).
The use of protein adducts as biomarkers also provided a
relative long lifespan with well-known kinetics such as in the
case of sulfur mustard (Benschop et al., 1997), acrylamide
(Hagmar et al., 2001), ethylene oxide (Tates et al., 1995),
dichlorvos (Mason, 2000), and acrylonitrile (Bader and
Wrbitzky, 2006).
After two incidents in Japan involving Sarin, samples were
collected within a few hours. Some of these samples were not
analyzed until several months later because the necessary
analytical techniques were yet to be developed (Minami et al.,
1997; Polhuijs et al., 1997; Noort et al., 1998; Fidder et al.,
2002). The results of these delayed chemical analyses
confirmed the suspicion that the clinical symptoms in the
victims were actually caused by Sarin exposure.
Successful quantitative reconstruction of exposure levels after
an incident have been reported, based on the measurement of
protein adducts (Tates et al., 1995; Bader and Wrbitzky, 2006)
and on the measurements of reaction products to enzymes in
plasma and erythrocytes (Mason, 2000). In Table 4, an overview
of the success factors based on past performance is given.
Complicating Factors
In the literature, factors were reported that delayed or
prevented the use of a systemic and standardized approach to
analysis of body tissues (see Supplementary Information,
255
Scheepers et al.
Table 4). The most important obstacles were related to
recruitment of study subjects and technical problems such as
contamination and loss of samples during transport and
storage (Ackermann-Liebrich et al., 1992), and formal
restrictions such as the lack of permission for analysis of
body tissues (Ackermann-Liebrich et al., 1992), delay in
sample collection because of too long discussions about the
necessity of biolocial monitoring (Traupe et al., 1997), and
lack of funds (Dayal et al., 1992). Many studies reported
limitations with respect to the involvement of study subjects
(Slottje et al., 2005a, b) or a too small population size (PerezCadahia et al., 2007).
Stepwise Procedure for Decision Making
A decision tree is considered useful not only in preparing
for biological monitoring campaigns after an incidental
chemical release, but also in the decision-making process.
Biological monitoring can be used to estimate the actual
exposure of a human population. However, this method
places significant demands on human and financial resources.
It is therefore desirable to apply it only if there is sufficient
certainty that people have been exposed to concentrations
that are high enough to cause adverse health effects and that
the biomarker will still be detectable at the time of sample
collection. The proposed decision tree has been developed to
meet these needs, and is illustrated by the acrylonitrile and
benzene examples. This tool will help to avoid sampling and
analyses of biological samples in cases in which biological
monitoring has no added value.
If AEGL values have not been established, this might
be a reason to refrain from the use of biological monitoring;
however, this method should be considered in high-exposure
situations with substances producing systemic effects and
judged by comparison with alternative threshold values such
as OELs (see Supplementary Information, Table 5).
Sampling Collection Strategy
For a quantitative analysis, it is of particular interest to
collect precise data as to the onset and the duration of the
exposure. Furthermore, it is recommended to perform
repeated measurements in time. For those biomarkers that
follow zero-order, or first-order kinetics, three or more
measurements support a calculation of the biomarker level
retrospectively, to the time point estimated as the end of the
chemical exposure after incident. More than three time points
of adduct determination will further improve the reliability of
the initial quantitative exposure estimate that is calculated by
extrapolation from a regression line.
Interpretation of Biomarker Levels
The biomarker level in a biological tissue (Ce) can be related
to the external concentration, based on relationships between
ambient inhalation exposure levels and biomarker levels that
have already been derived for biological limit values
256
Biomonitoring following chemical incidents
(ACGIH, 2009; DFG, 2008, see Supplementary Information, Table 5). If such relationships have not yet been
established, it is possible to use physiologically based
pharmacokinetic (PBPK) models to derive a calculated
estimate, especially for those substances for which such a
PBPK model was used to derive an AEGL (Bruckner et al.,
2004; Krewski et al., 2004; Boyes et al., 2005; Bos et al.,
2006). Clewell et al. (2008) have recently proposed a
procedure for a simple direct translation from biomarker
levels to ambient exposure estimates using exposure conversion factors (ECFs). These factors take into account the
biological variability within a population by using an ECF
distribution (Clewell et al., 2008). Using these calculations, it
is possible to predict whether biomarker levels correspond to
an exposure that exceeds an established IVER. As the
derivation of ECF distributions could be time consuming, it
would be necessary to establish these values for different
chemicals in advance.
Preparedness
If biological monitoring is considered a possible means of
exposure assessment before an incident, suitable arrangements should be made in advance. Below, some aspects of
preparedness are discussed.
1. Preparing a database with the key parameters relevant for
the decision tree. These data include updated information
on the half-lives of biomarkers in humans and on the
LOQ of analytical methods available for the determination of biomarkers in body tissues. In addition, IVERs are
usually easily available but for (groups of) substances for
which no IVERs are available, data sources should be
identified beforehand.
2. Materials, such as needles and containers for sample
collection, may be kept in stock at a central repository.
Containers used for collection of urine could be pretreated (e.g., acid rinsed) and kept in stock, ready for use.
3. Protocols for sample collection, transport, and storage
may be prepared for large groups of chemical substances
for which the same physicochemical and technical
conditions apply.
4. Questionnaires needed to support interpretation of outcomes of biological monitoring campaigns could be
compiled for large groups of chemicals.
5. Computer models such as PBPK models could be made
available for immediate use for reverse modeling of the
relationship of a biomarker value with the external
exposure (to evaluate if the exposure exceeded an
intervention level at the time of the incident). ECF
distributions could be prepared in advance.
Risk Management
If biological materials can be analyzed in an early stage after
the incident, the outcome of these measurements may be used
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
in decisions concerning the medical treatment of individuals.
In individuals who suffer from health effects, such data may
confirm the identity of a specific substance that can then be
considered as a possible cause and the severity of the signs of
intoxication may be related to the biomarker level, for
example, in the case of neurological symptoms (Hagmar
et al., 2001) or inhibition of different types of cholinesterase
activity in plasma or erythrocytes after exposure to an
organophosphor ester insecticide, carbamate pesticide, or to
a nerve agent (Benschop et al., 1997; Polhuijs et al., 1997;
Tsuchihashi et al., 1998; Mason, 2000; Fidder et al., 2002). If
the results of biological monitoring become available in a
later stage, that is, after victims have died or survivors have
completely recovered, these results are not relevant to the
medical treatment of individuals (Tates et al., 1995; Alarie,
2002; Clewell et al., 2008). However, in these cases, these
obtained data may still have a role in improvement of first
response to chemical incidents and medical support of victims
suffering from intoxications in the future.
Scheepers et al.
recommended that, considering the potential of biological
monitoring to yield valuable data on exposure, every
opportunity should be taken to obtain blood and urine
specimens from exposed workers and members of the
affected population (WHO, 1997, 2009).
Many aspects need to be taken into account when deciding
about the usefulness and necessity of implementing biological
monitoring after a chemical incident. These include the
kinetics pattern of elimination of the biomarker, the toxicity
mechanism(s) involved, the chemical analytical possibilities
for determining levels of the biomarker in biological tissues,
and knowledge about background values of a biomarker in
the general population. The stepwise approach using the
proposed decision tree is considered to be a useful tool in the
decision-making process after an acute chemical incident.
Conflict of interest
The authors declare no conflict of interests.
Risk Communication
It is important to relay the message to the victims that results
of a biological monitoring campaign will be reported
anonymously on a group basis (Scheepers, 2009). Individual
results will usually only be communicated with the physician
treating the victim, for example, in the case of clinically
relevant deviations from expected baseline values.
In many cases, results of biological monitoring will show
that exposure to hazardous substances because of a chemical
incident has been below levels expected to induce adverse
health effects. Such findings may be important to report to
the victims of such an incident and may prevent attribution of
health complaints to a physical factor (Ackermann-Liebrich
et al., 1992; Edelman et al., 2003; Roorda et al., 2004; Wolff
et al., 2005).
In case it is decided not to start a biological monitoring
campaign, this should be clearly communicated to the
population involved in the incident. It is also important that
the reasons for not proceeding with biological monitoring
should be communicated appropriately and convincingly on a
societal level. Information retrieved from the stepwise procedure
for decision making can be used in this communication.
Conclusions
On a societal level, a disaster might cause unrest and mistrust
of the government. On this level, health studies may also
serve as a signal of recognition of the problems of victims.
Survivors of disasters need to know that their ideas about
exposure to toxic substances are being taken seriously (Van
den Berg et al., 2008). Biological monitoring as part of health
studies is an intervention that might help the survivors to
master control over their own situation. It has been
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Acknowledgements
This study was funded by the National Institute for Pubic
Health and the Environment, The Netherlands. We express
our gratitude to Dr. Paul Aston of AB Biomonitoring for
his suggestions to improve the language of the manuscript.
We are indebted to Dr. Peter Heitland and Dr. HansWolfgang Hoppe of the Medizinisches Labor Bremen for
supplying technical data concerning the analysis of biomarkers in body fluids.
References
ACGIH. Documentation of the Biological Exposure Indices, 7th edn. ACGIH,
Cincinnati, OH, 2001a.
ACGIH. Xylenes (Technical or Commercial Grades). BEI Documentation.
ACGIH, Cincinnati, OH, 2001b.
ACGIH. TLVs and BEIs. ACGIH, Cincinnati, OH, 2009.
Ackermann-Liebrich U.A., Braun C., and Rapp R.C. Epidemiologic analysis of an
environmental disaster: the Schweizerhalle experience. Environ Res 1992: 58: 1–14.
Ahn H.S., and Shin H.S. Determination of ethylene oxide-hemoglobin adduct by
silylation and gas chromatography-electron impact-mass spectrometry.
J Chromatogr 2006: 843: 202–208.
Alarie Y. Toxicity of fire smoke. Crit Rev Toxicol 2002: 32: 259–289.
Allain P., Mauras Y., Dougé C., Jaunault L., Delaporte T., and Beaugrand C.
Determination of iodine and bromine in plasma and urine by inductively
coupled plasma mass spectrometry. Analyst 1990: 115: 813–815.
Angerer J., and Kramer A. Occupational chronic exposure to organic solvents.
XVI. Ambient and biological monitoring of workers exposed to toluene. Int
Arch Occup Environ Health 1997: 69: 91–96.
Angerer J, and Schaller K.H Analyses of hazardous substances in biological
materials. Deutsche Forschungsgemeinschaft VCH, Weinheim, Vol. 4, 1985,
pp 107–130.
ATSDR. Toxicological profile for styrene. 2007. Available: http://www.atsdr.cdc.
gov/toxprofiles/tp53.html (accessed February 8, 2010).
ATSDR. Toxicological profile for bromomethane. 1992. Available: http://www.
atsdr.cdc.gov/toxprofiles/tp27.html (accessed February 8, 2010).
257
Scheepers et al.
Aubin J., Potvin L., Beland F., and Pineault R. Utilisation of a specialized clinic
following an ecological accident. Med Care 1994: 32: 1–14.
Bader M., and Wrbitzky R. Follow-up biological monitoring after accidental
exposure to acrylonitrileFimplications for protein adducts as a dose monitor
for short-term exposures. Toxicol Lett 2006: 162: 125–131.
Baselt R.C. Disposition of Toxic Drugs and Chemicals in Man, 5th edn. Chemical
Toxicology Institute, 2000.
Baud F.J., Borron S.W., Bavoux E., and Astier A., Hoffman J.R. Relation
between plasma lactate and blood cyanide concentrations in acute cyanide
poisoning. Brit Med J 1996: 312: 26–27.
Benschop H.P., van der Schans G.P., Noort D., Fidder A., Mars-Groenendijk
R.H., and de Jong L.P. Verification of exposure to sulfur mustard in two
casualties of the Iran-Iraq conflict. J Anal Toxicol 1997: 21: 249–251.
Bergmark E. Hemoglobin adducts of acrylamide and acrylonitrile in laboratory
workers, smokers and nonsmokers. Chem Res Toxicol 1997: 10: 78–84.
Bertazzi P.A. Long-term effects of chemical disasters. Lessons and results from
Seveso. Sci Total Environ 1991: 106: 5–20.
Bertazzi P.A., Bernucci I., Brambilla G., Consonni D., and Pesatori A.C. The
Seveso studies on early long-term effects of dioxin exposure: a review. Environ
Health Perspect 1998: 106(Suppl. 2): 625–633.
Bongers S., Janssen N.A.H., Reiss B., Grievink L., Lebret E., and Kromhout H.
Challenges of exposure assessment for health studies in the aftermath of chemical
incidents and disasters. J Exposure Science Environ Epid 2008: 18: 341–359.
Boogaard P.J., and van Sittert N.J. Biological monitoring of exposure to benzene:
a comparison between S-phenylmercapturic acid, trans,trans-muconic acid,
and phenol. Occup Environ Med 1995: 52: 611–620.
Bos P.M.J., Zeilmaker M.J., and van Eijkeren J.C.H. Application of
physiologically based pharmacokinetic modeling in setting acute exposure
guideline levels for methylene chloride. Toxicol Sci 2006: 91: 576–585.
Boyes W.K., Evans M.V., Eklund C., Janssen P., and Simmons J.E. Duration
adjustment of acute exposure guideline level values for trichloroethylene using
a physiologically-based pharmacokinetic model. Risk Analysis 2005: 25:
677–686.
Bruckner J.V., Keys D.A., and Fisher J.W. The acute exposure guideline level
(AEGL) program: applications of physiologically based pharmacokinetic
modeling. J Toxicol Environ Health A 2004: 67: 621–634.
Brugnone F., Perbellini L., Apostoli P., Locatelli M., and Mariotto P. Decline of
blood and alveolar toluene concentration following two accidental human
poisonings. Int Arch Occup Environ Health 1983: 53: 157–165.
Carmella S.G., Chen M., Zhang Y., Zhang S., Hatsukami D.K., and Hecht S.S.
Quantitation of acrolein-derived (3-hydroxypropyl)mercapturic acid in human
urine by liquid chromatography-atmospheric pressure chemical ionization
tandem mass spectrometry: effects of cigarette smoking. Chem Res Toxicol
2007: 20: 986–990.
Chambers D.M., McElprang D.O., Waterhouse M.G., and Blount B.C. An
improved approach for accurate quantitation of benzene, toluene, ethylbenzene, xylene, and styrene in blood. Anal Chem 2006: 78: 5375–5383.
Clewell H.J., Tan Y.M., Campbell J.L., and Andersen M.E. Quantitative
interpretation of human biological monitoring data. Toxicol Appl Pharmacol
2008: 231: 122–133.
Dayal H.H., Brodwick M., Morris R., Baranowski T., Trieff N., and Harrison
J.A. et al. A community-based epidemiologic study of health sequelae of
exposure to hydrofluoric acid. Ann Epidemiol 1992: 2: 213–230.
DFG. List of MAK and BAT Values. Commission for the Investigation of Health
Hazards of Chemical Compounds in the Work Area. DFG Report No. 44
Wiley-VCH, Weinheim, 2008.
Edelman P., Osterloh J., Pirkle J., Caudill S.P., Grainger J., and Jones R. et al.
Biomonitoring of chemical exposure among New York City firefighters
responding to the World Trade Center fire and collapse. Environ Health
Perspect 2003: 111: 1906–1911.
Engström K., and Riihimäki V. Determination of methylhippuric acids in urine by
gas chromatography. IARC Sci Publ 1988: 85: 313–318.
Ekstrand J., and Ehrnebo M. The relationship between plasma fluoride, urinary
excretion rate and urine fluoride concentration in man. J Occup Med 1983: 25:
745–748.
EPA. Acute exposure guideline levels for acyrylonitrile, 2007. Available:
http://www.epa.gov/oppt/aegl/pubs/acrylonitrile_proposed_jul_2007.v1.pdf
(accessed July 10, 2009).
EPA. Interim acute exposure guidelines level for benzene. NAS/COT Subcommittee
for benzene, 2009a. Available: http://www.epa.gov/oppt/aegl/pubs/benzene_
interim_dec_2008_v1.pdf (accessed July 10, 2009).
258
Biomonitoring following chemical incidents
EPA. CAMEO: Downloading, installing, and running ALOHA, 2009b. Available:
http://www.epa.gov/OEM/content/cameo/aloha.htm (accessed December 1,
2009).
Fidder A., Hulst A.G., Noort D., de Ruiter R., van der Schans M.J., and
Benschop H.P., et al. Retrospective detection of exposure to organophosphorus anti-cholinesterases: mass spectrometric analysis of phosphylated
human butyrylcholinesterases. Chem Res Toxicol 2002: 15: 582–590.
Geyer H.J., Schramm K.W., Feicht E.A., Behechti A., Steinberg C., and
Bruggemann R., et al. Half-lives of tetra-, penta-, hexa-, hepta-, and
octachlorodibenzo-p-dioxin in rats, monkeys, and humans. A critical review.
Chemosphere 2002: 48: 631–644.
Ghittori S., Maestri L., Florentino M.L., and Imbriani M. Evaluation of
occupational exposure to benzene by urine analysis. Int Arch Occup Environ
Health 1995: 67: 195–200.
Hagmar L., Törnqvist M., Nordander C., Rosén I., Bruze M., and Kautiainen A.
Health effects of occupational exposure to acrylamide using hemoglobin
adducts as biomarkers of internal dose. Scand J Work Environ Health 2001:
27: 219–226.
Hasegawa K., Shiojima S., Koizumi A., and Ikeda M. Hippuric acid and o-cresol
in the urine of workers exposed to toluene. Int Arch Occup Environ Health
1983: 52: 197–208.
Haufroid V., Merz B., Hofmann A., Tschopp A., Lison D., and Hotz P. Exposure
to ethylene oxide in hospitals: biological monitoring and influence of
glutathione S-transferase and epoxide hydrolase polymorphisms. Cancer
Epidemiol Biomarkers Prev 2007: 16: 796–802.
Heitland P., and Köster H.D. Biomonitoring of 30 trace elements in urine of
children and adults by ICP-MS. Clinica Chimica Acta 2006: 365: 310–318.
Heitland P., and Köster H.D. Comparison of different medical cases in urinary
arsenic speciation by fast HPLC-ICP-MS. Int J Hyg Environ Health 2008:
212: 432–438.
Itai K., and Tsunoda H. Highly sensitive and rapid method for determination
of fluoride ion concentrations in serum and urine using flow injection
analysis with a fluoride ion-selective electrode. Clin Chim Acta 2001: 308:
163–171.
Jakubowski M., Linhart I., Pielas G., and Kopecky J. 2-Cyanoethylmercapturic
acid (CEMA) in the urine as a possible indicator of exposure to acrylonitrile.
Br J Ind Med 1987: 44: 834–840.
Jarup L., Rogenfelt A., Elinder C.G., Nogawa K., and Kjellstrom T. Biological
half-time of cadmium in the blood of workers after cessation of exposure.
Scand J Work Environ Health 1983: 9: 327–331.
Jongeneelen F.J., Anzion R.B., and Henderson P.T. Determination of hydroxylated
metabolites of polycyclic aromatic hydrocarbons in urine. J Chromatogr 1987:
413: 227–232.
Jongeneelen F.J., Anzion R.B., Scheepers P.T., Bos R.P., Henderson P.T., and
Nijenhuis E.H., et al. 1-Hydroxypyrene in urine as a biological indicator of
exposure to polycyclic aromatic hydrocarbons in several work environments.
Ann Occup Hyg 1988: 32: 35–43.
Kataoka H., Manabe K., Nakase S., and Makita M. Determination of hippuric
acid and o-, m- and p-methylhippuric acids in urine by capillary gas
chromatography. J Pharm Biomed Anal 1991: 9: 699–704.
Kawai T., Okada Y., Odachi T., Horiguchi S., Zhang Z., and Moon C., et al.
Diffusive sampling and biological monitoring of 2-bromopropane. Arch
Environ Contam Toxicol 1997: 33: 23–28.
Jakasa I., and Wenker M. Determination of mandelic acid enantiomers in urine by
gas chromatography and electron-capture or flame ionisation detection.
J Chromatogr 2000: 738: 39–46.
Krewski D., Bakshi K., Garrett R., Falke E., Rusch G., and Gaylor D.
Development of acute exposure guideline levels for airborne exposures to
hazardous substances. Reg Toxicol Pharmacol 2004: 39: 184–201.
Lauwerys R., and Hoet P. Industrial Chemical Exposure, Guidelines for Biological
Monitoring. Lewis, Boca Raton, 2001.
Leung H.W., Kerger B.D., and Paustenbach D.J. Elimination half-lives of selected
polychlorinated dibenzodioxins and dibenzofurans in breast-fed human
infants. J Toxicol Environ Health A 2006: 69: 437–443.
Lewalter J., Korallus U., Harzdorf C., and Weidemann H. Chromium bond
detection in isolated erythrocytes: a new principle of biological monitoring of
exposure to hexavalent chromium. Int Arch Occup Environ Health 1985: 55:
305–318.
Mason H.J. The recovery of plasma cholinesterase and erythrocyte acetylcholinesterase activity in workers after over-exposure to dichlorvos. Occup Med
2000: 50: 343–347.
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
Minami M., Hui D.M., katsumata M., Inagaki H., and Boulet C.A. Method for
the analysis of the methylphosphonic acid metabolites of sarin and its ethanolsubstituted analogue in urine as applied to the victims of the Tokyo sarin
disaster. J Chromatogr B 1997: 695: 237–244.
McKelvey W., Gwynn R.C., Jeffery N., Kass D., Thorpe L.E., and Garg R.K.,
et al. A biomonitoring study of lead, cadmium, and mercury in the blood of
New York city adults. Environ Health Perspect 2007: 115: 1435–1441.
Michigami Y., Yamamoto Y., and Ueda K. Determination of nitrite, sulphate,
bromide and nitrate in human serum by ion chromatography. Analyst 1989:
114: 1201–1205.
Müler A.M.F., Hallier E., Westphal G., Schröder K.R., and Bolt H.M.
Determination of methylated globin and albumin for biomonitoring of
exposure to methylating agents using HPLC with precolumn fluorescent
derivatization. Fresenius J Anal Chem 1994: 350: 712–715.
Needham L.L., Gerthoux P.M., Patterson Jr D.G., Brambilla P., Smith S.J., and
Sampson E.J., et al. Exposure assessment: serum levels of TCDD in Seveso,
Italy. Environ Res 1999: 80: S200–S206.
Nomiyama K., and Nomiyama H. Respiratory elimination of organic solvents in
man. Benzene, toluene, n-hexane, trichloroethylene, acetone, ethyl acetate and
ethyl alcohol. Int Arch Arbeitsmed 1974: 32: 85–91.
Noort D., Hulst A.G., Platenburg D.H., Polhuijs M., and Benschop H.P.
Quantitative analysis of O-isopropyl methylphosphonic acid in serum samples
of Japanese citizens allegedly exposed to sarin: estimation of internal dosage.
Arch Toxicol 1998: 72: 671–675.
National Research Council. Standing Operating Procedures for Developing Acute
Exposure Guideline Levels for Hazardous Chemicals. Subcommittee on Acute
Exposure Guideline Levels, Committee on Toxicology, National Academy
Press, Washington, D.C., 2001.
Odoul M., Fouillet B., Nouri B., Chambon R., and Chambon P. Specific
determination of cyanide in blood by headspace gas chromatography. J Anal
Toxicol 1994: 18: 205–207.
Perez-Cadahia B., Lafuente A., Cabaleiro T., Pasaro E., Mendez J., and
Laffon B. Initial study on the effects of Prestige oil on human health. Environ
Int 2007: 33: 176–185.
Petersen R., Thomsen J.F., Jrgensen N.K., and Mikkelsen S. Half life of
chromium in serum and urine in a former plasma cutter of stainless steel.
Occup Environ Med 2000: 57: 140–142.
Pierre F., Baruthio F., Diebold F., and Biette P. Effect of different exposure
compounds on urinary kinetics of aluminium and fluoride in industrially
exposed workers. Occup Environ Med 1995: 52: 396–403.
Polhuijs M., Langenberg J.P., and Benschop H.P. New method for retrospective
detection of exposure to organophsophorus anticholinesterases: application
of alleged sarin victims of Japanese terrorists. Tox Appl Pharm 1997: 146:
156–161.
Pomroy C., Charbonneau S.M., McCullough R.S., and Tam G.K. Human
retention studies with 74As. Toxicol Appl Pharmacol 1980: 53: 550–556.
Quinones O., Snyder S.A., Cotruvo J.A., and Fisher J.W. Analysis of bromate
and bromide in blood. Toxicology 2006: 221: 229–234.
Rauws A.G. Pharmacokinetics of bromide ion F an overview. Food Chem
Toxicol 1983: 21: 379–382.
Roorda J., van Stiphout W.A., and Huijsman-Rubingh R.R. Post-disaster health
effects: strategies for investigation and data collection. Experiences from the
Enschede firework disaster. J Epidemiol Community Health 2004: 58: 982–987.
Ryan M., and Baumann R.J. Use and monitoring of bromides in epilepsy
treatment. Pediatr Neurol 1999: 21: 523–528.
Rueff J., Teixeira J.P., Santos L.S., and Gaspar J.F. Genetic effects and biotoxicity
monitoring of occupational styrene exposure. Clinica Cimica Acta 2009: 399:
8–23.
Sangster B., Blom J.L., Sekhuis V.M., Loeber J.G., Rauws A.G., and Koedam
J.C., et al. The influence of sodium bromide in man: a study in human
volunteers with special emphasis on the endocrine and the central nervous
system. Food Chem Toxicol 1983: 21: 409–419.
Schaller K.H., Csanady G., Filser J., Jungert B., and Drexler H. Elimination
kinetics of metals after an accidental exposure to welding fumes. Int Arch
Occup Environ Health 2007: 80: 635–641.
Schecter A., Ryan J.J., and Päpke O. Decrease in levels and body burden of
dioxins, dibenzofurans, PCBS, DDE, and HCB in blood and milk in a mother
nursing twins over a thirty-eight month period. Chemosphere 1998: 37: 1807–1816.
Scheepers P. Biomarkers of exposure to carcinogens. In: Ballantyne B., Marrs
T.C., and Syversen T. (Eds.). General and Applied Toxicology, 3rd edn. John
Wiley and Sons, Chichester, West Sussex, UK, 2009, pp. 1841–1855.
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Scheepers et al.
Schettgen T., Musiol A., and Kraus T. Simultaneous determination of mercapturic
acids derived from ethylene oxide (HEMA), propylene oxide (2-HPMA),
acrolein (3-HPMA), acrylamide (AAMA) and N,N-dimethylformamide
(AMCC) in human urine using liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom 2006: 22: 2629–2638.
Schettgen T., Musiol A., Alt A., Ochsmann E., and Kraus T. A method for the
quantification of biomarkers of exposure to acrylonitrile and 1,3-butadiene in
human urine by column-switching liquid chromatography-tandem mass
spectrometry. Anal Bioanal Chem 2009: 393: 969–981.
Shibata M., Inoue K., Yoshimura Y., Nakazawa H., and Seto Y. Simultaneous
determination of hydrogen cyanide and volatile aliphatic nitriles by headspace
gas chromatography, and its application to an in vivo study of the metabolism
of acrylonitrile in the rat. Arch Toxicol 2004: 78: 301–305.
Shiono H., Maseda Ch., Akane A., and Matsubara K. Rapid and sensitive
quantitation of cyanide in blood and its application to fire victims. Am J Foren
Med Pathol 1991: 12: 50–53.
Schulz V., Bonn R., and Kindler J. Kinetics of elimination of thiocyanate in
7 healthy subjects and in 8 subjects with renal failure. Klin Wochenschr 1979:
57: 243–247.
Slottje P., Bijlsma J.A., Smidt N., Twisk J.W., Huizink A., and Lems W.F., et al.
Epidemiological study of the autoimmune health effects of a cargo aircraft
disaster. Arch Intern Med 2005a: 165: 2278–2285.
Slottje P., Huizink A.C., Twisk J.W., Witteveen A.B., van der Ploeg H.M., and
Bramen I., et al. Epidemiological study air disaster in Amsterdam (ESADA):
study design. BMC Public Health 2005b: 5: 54.
Stanley J.S., Onstot J., and Sack T. PCDDs and PCDFs in human adipose tissue
from the EPA FY 82 NHATS repository. Chemosphere 1986: 15: 1605–1612.
Symanski E., Bergamaschi E., and Mutti A. Inter- and intra-individual sources of
variation in levels of urinary styrene metabolites. Int Arch Occup Environ
Health 2001: 74: 336–344.
Tates A.D., Boogaard P.J., Darroudi F., Natarajan A.T., Caubo M.E., and van
Sittert N.J. Biological effect monitoring in industrial workers following
incidental exposure to high concentrations of ethylene oxide. Mutat Res 1995:
329: 63–77.
Teixeira J.P., Silva S., Torres J., Gaspar J., Roach H., Farmer P.B., Rueff J., and
Mayan O. Styrene-oxide N-terminal valine haemoglobin adducts as biomarkers of occupational exposure to styrene. Int J Hyg Enviorn Health 2008:
211: 59–62.
Törnqvist M., Fred C., Haglund J., Helleberg H., Paulsson B., and Rydberg P.
Protein adducts: quantitative and qualitative aspects of their formation,
analysis and applications. J Chromatogr B 2002: 778: 279–308.
Trainor M.T., Ridgway P., Macbeth R.W., Wilday A.J., and Balmforth H.F.
Substance priotisation for the development of EU acute exposure toxicity
thresholds (AETLs). J Hazardous Materials 2006: 133: 13–23.
Traupe H., Menge G., Kandt I., and Karnaus W. Higher frequency of atopic
dermatitis and decrease in viral warts among children exposed to chemicals
liberated in a chemical accident in Frankfurt, Germany. Dermatology 1997:
195: 112–118.
Tsuchihashi H., Katagi M., Nishikawa M., and Tatsuno M. Identification of
metabolites of nerve agent VX in serum collected from a victim. J Anal Toxicol
1998: 22: 383–388.
Van den Berg B., Grievink L., Gutschmidt K., Lang T., Palmer S., and
Ruijten M. et al. The public health dimension of disasters-health outcome
assessment of disasters. Prehospital Disast Med 2008: 23: s55–s59.
Walker V.E., MacNeela J.P., Swenberg J.A., Turner M.J. Jr., and Fennell T.R.
Molecular dosimetry of ethylene oxide: formation and persistence of
N-(2-hydroxyethyl)valine in hemoglobin following repeated exposures of
rats and mice. Cancer Res 1992: 15: 4320–4327.
Weber L. Gas chromatographic determination of urinary phenol conjugates after
acid hydrolysis/extractive acetylation. J Chromatogr 1992: 574: 349–351.
WHO. Manual for the Public Health Management of Chemical Incidents. WHO,
Geneva, 2009.
WHO Regional Office for Europe. Assessing the health consequences of major
chemical incidentsFepidemiological approaches. WHO Reg Publ Eur Ser
1997: 79: 1–90.
Wisner B., and Adams J. Environmental Health in Emergencies and Disasters, a
Practical Guide. WHO, Geneva, 2002.
Wolff M.S., Teitelbaum S.L., Lioy P.J., Santella R.M., Wang R.Y., Jones R.L.,
Caldwell K.L., Sjödin A., Turner W.E., Li W., Georgopoulos P., and Berkowitz
G.S. Exposures among pregnant women near the World Trade Center site on 11
September 2001. Environ Health Perspect 2005: 113: 739–748.
259
Biomonitoring following chemical incidents
Scheepers et al.
Appendix A
In this appendix we will describe how the elapsed time ts
between the end of exposure during an incident and the last
possible moment for sample collection can be calculated. We
will make a distinction between two situations: first-order
kinetics as typical for the elimination from body tissues of
most parent compounds and their metabolites. For those
biomarker that are captured in blood cells, such as
hemoglobin adducts, we will also describe how to perform
these calculations for biomarkers that follow zero-order
kinetics. For both types of biomarkers, an example will
be given.
Cts, which is defined as the concentration at the time of
collection of biological tissues, should be equal to or above
the limit of quantification:
Cts LOQ
ðA:1Þ
LOQ indicates the minimum concentration of the biomarker at the time of sampling. It may be replaced by an
alternative criterion for positive identification or by the
95 percentile background of the biomarker level in the
general population if this value is substantially higher than
the LOQ.
First-order elimination
When the biomarker is assumed to follow a log-linear decline
in time, the value of this parameter at any time point (Cts )
after cessation of the exposure can be described as:
C ts ¼
2
Ce
ðts =t1=2 Þ
ðA:2Þ
In which Ce is the concentration of the biomarker at the end
of the exposure and t1/2; is the half-life of this parameter
(in hours), and ts is the time lapsed between the end of the
exposure at the chemical incident location and the time when
biological materials can be collected. Equations (1) and (2)
can be rewritten as:
Ce LOQ 2ðts =t1=2 Þ
ðA:3Þ
concluded that the excretion of S-phenyl mercapturic acid
(SPMA) is the most sensitive biomarker of exposure in urine
and reported a t1/2; of 9.0±4.5 h or urinary elimination of
SPMA. Further, the following regression equation was
reported for an 8-h TWA occupational exposure (DFG,
2008):
½SPMAurine ðmg=g creatinineÞ ¼ 0:045 ½benzeneair ðp:p:m:Þ
0:001 ðr ¼ 0:999Þ
ðA:4Þ
Using this equation an exposure to the 8-h AEGL-2 of
200 p.p.m. corresponds to an SPMA concentration of
E9,000 mg/g creatinine. For the purpose of conversion,
1 mg/g creatinine is assumed to be equal to 1 mg/l. With a
95 percentile value (P95) for the background of SPMA in
urine of 7.3 mg/l (Scheepers, 2009), the critical longest postexposure sampling time (ts) can be calculated as:
ts ¼
t1=2
Ce
logð Þ
0:30
P95
Completing Eq. (5) with the data for benzene gives ts E93 h.
This means that urine should be collected within
approximately 4 days after the incident. Assuming that the
estimated exposure during the incident described in the
case study was approximately 300 p.p.m. for 8 h, not
much more time is available to organize urine sampling
(ts E98 h).
It is noted that different relationships between benzene
concentrations in air and urinary SPMA concentrations have
been published (ACGIH, 2001). For instance, based on the
relationship published by Ghittori et al. (1995) an air
concentration of 200 p.p.m. benzene would lead to a
calculated SPMA concentration of 2,006 mg/g creatinine,
which would give an estimation of tsE73 h (3 days).
Zero-order elimination
For those biomarkers that follow zero-order kinetics, such as
protein adducts encapsulated in circulating blood cells,
another equation should be used:
Cts ¼ a Ce t þ Ce
The second part of this equation (2ðts =t1=2 Þ ) indicates the factor
by which the concentration at the end of exposure has
decreased as a function of the number of half-lives that
elapsed between the end of the exposure and the moment of
sampling ts.
Case 1: Benzene
The example of benzene is elaborated, based on the
publication by Boogaard and Van Sittert (1995). They
260
ðA:5Þ
ðA:6Þ
In this equation, a is the slope of the function describing the
loss of adduct per day and is dependent on the lifespan of
hemoglobin, which is equal to the lifespan of the erythrocyte
(ter; Törnqvist et al., 2002). This lifespan is 126 days, and
hence:
a¼
1
0:008
ter
ðA:7Þ
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
Biomonitoring following chemical incidents
Scheepers et al.
If the adduct is stable, the lifespan of the erythrocyte is equal
to the lifespan of the adduct. The lifespan of the adduct is
twofold the biomarker half-life:
a¼
1
2t1=2
ðA:8Þ
For t ¼ ts, Eq. (6) can be rewritten as:
ts ¼
Ce Cts
aCe
ðA:9Þ
Using Eq. (8) this is equivalent to:
C e C ts
a Ce
Case 2: acrylonitrile
For acrylonitrile the parameter estimates are taken from an
incident described by Bader and Wrbitzky (2006). They
reported an effective half-life for the adduct of B75 days in
humans. The researchers suggest that this half-life is longer than
the half-life of hemoglobin (63 days) because of a depot of
circulating reactive intermediates that leads to an extended
internal exposure relative to the duration of external exposure.
For the case study of the spill of acrylonitrile in Amersfoort, it
is assumed that exposure occurs at or around the 1-h AEGL-2
of 58 p.p.m. An established linear relationship between the
ambient air concentration and the level of globine adducts
(DFG, 2008) is used to estimate the cyanoethylvaline adduct
level by extrapolation from this relationship with linear equation:
ðA:10Þ
½Cyanoethylvalineblood ðmg=lÞ ¼ 140:1 ½acrylonitrateair ðp:p:m:Þ
1:360 ðr ¼ 0:999Þ
At t ¼ ts, Cts becomes equal to LOQ (or P95 if the
background level of the biomarker is higher than
the LOQ). If LOQ (or P95) 5 Ce Eq. (10) can be
simplified to:
ðA:12Þ
ts ¼ 2 t1=2
ts ffi 2 t1=2
ðA:11Þ
In chemical incidents this simplification can often be
used because of the extremely high exposure relative
to the value of the LOQ or the background biomarker
level.
This yields an adduct level Ce of 8,142 mg/l blood. With a
value of LOQ of 0.05 mg/l, Eq. (10) can be used as 0.05 5
8,142. This does not change even if a background of this
adduct is assumed in smokers that exceeds the LOQ by a
factor of 40 (Scheepers, 2009).
ts ffi 2:75
ðA:13Þ
The value for ts is estimated to be 150 days.
Supplementary Information accompanies the paper on the Journal of Exposure Science and Environmental Epidemiology
website (http://www.nature.com/jes)
Journal of Exposure Science and Environmental Epidemiology (2011) 21(3)
261