Framework for the development and application of environmental biological monitoring guidance values

Regulatory Toxicology and Pharmacology 63 (2012) 453–460

Contents lists available at SciVerse ScienceDirect

Regulatory Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/yrtph

Framework for the development and application of environmental biological monitoring guidance values Ruth Bevan a,⇑, Juergen Angerer b, John Cocker c, Kate Jones c, Holger M. Koch b, Ovnair Sepai d, Greet Schoeters e, Roel Smolders e, Len Levy a a

Institute of Environment and Health, Cranfield Health, Cranfield University, Bedfordshire MK43 0AL, UK Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Ruhr University Bochum, Germany Health and Safety Laboratory, Harper Hill, Buxton, Derbyshire SK17 9JN, UK d Health Protection Agency, Chilton, Didcot, Oxon OX11 0RQ, UK e Vito, Boeretang 200, BE-2400 MOL, Belgium b c

a r t i c l e

i n f o

Article history: Received 7 December 2011 Available online 7 June 2012 Keywords: Biomonitoring Biological guidance values Framework Environmental exposure Chemicals Risk assessment Biological monitoring Exposure assessment Risk communication Interpretation

a b s t r a c t Human biomonitoring (HBM) is widely recognised as a useful tool to aid assessment of exposure to chemical substances, but our ability to detect hazardous substances (or their metabolites and health effects) often exceeds our understanding of their biological relevance. There are only a few established frameworks for developing and using occupational and environmental biological guidance values (BGVs), mostly for data-rich substances that have been in use for some time. BGVs for new substances and those with unknown dose–response relationships are difficult to derive. An accepted framework based on current scientific knowledge and best practice is therefore urgently needed to help scientists, regulators, and stakeholders to design appropriate HBM studies, interpret HBM data (both for groups and individuals) understand the limitations and to take appropriate action when required. The development and application of such a tool is described here. We derived a conceptual framework that was refined by consultation with an advisory group and workshop. The resulting framework comprised four levels defined by increasing data, with increasing confidence for human health risk assessment. Available data were used for 12 chemicals with expert judgement to illustrate the utility of the framework. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction The measurement of chemicals in human bodily fluids has traditionally been used to assess occupational exposures in workers, where correlations between biomonitoring values and personal air monitoring data, or early and reversible biological effects have been used to set biological guidance values (estimated levels of exposure associated with no adverse health effects) for a number of substances. In the 1970s, application of biomonitoring to environmental exposures facilitated the discovery of harmful levels of lead in the environment and resulted in the recognition of the potential of human biomonitoring (HBM) for monitoring environmental exposure of the general population to pollutants. Since that time, several national and international bodies (Academy of Sciences of the USA http://www.nap.edu/ openbook.php?record_id=11700&page=R1; Health and Environmental Science Institute of the International Life Science Institute ⇑ Corresponding author. Fax: +44 (0)1234 758280. E-mail address: r.bevan@cranfield.ac.uk (R. Bevan). 0273-2300/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yrtph.2012.06.002

http://www.hesiglobal.org/i4a/pages/index.cfm?pageID=3465; Deutsche Forschungsgemeinschaft http://www.dfg.de/en/dfg_profile/statutory_bodies/senate/health_hazards/structure/working_ groups/threshold_biological/index.html), have demonstrated the utility of HBM as a valuable public health tool. HBM provides a measure of the internal dose and is reflective of uptake from all routes of exposure. It can also be used to assess trends and changes in exposure of a population to environmental chemicals, and to help identify susceptible populations, such as children, who may be at higher risk from exposure. Current HBM initiatives in Europe include the German Environmental Survey (http://www.umweltbundesamt.de/gesundheit-e/ survey/index.htm), the Flemish human biomonitoring programme (http://www.milieu-en-gezondheid.be/index.html), Polish biomonitoring programmes (Indulski et al., 1999; Heinrich-Ramm et al., 2000; Jakubowski and Trzcinka-Ochocka, 2005) and an EU wide project, COPHES (Consortium to perform human biomonitoring on a European scale; http://www.eu-hbm.info/cophes). There are several ongoing initiatives to assess exposure of the US population to

454

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

a number of environmental chemicals, with the most prominent biomonitoring programme being the National Health and Nutrition Examination Survey (NHANES). In addition, the US EPA has conducted several other field-monitoring studies which include the National Human Exposure Assessment Survey (http://www.epa.gov/ nerl/research/nhexas/nhexas.htm/); Agricultural Health Study/Pesticide Exposure Study (http://www.aghealth.org/); Children’s Total Exposure to Persistent Pesticides and Other Persistent Organic Pollutants (http://www.epa.gov/heasd/ctepp/index.html/, Wilson et al., 2004); the CDC report on human exposure to environmental chemicals (http://www.cdc.gov/exposurereport/). In Canada, the Canadian Health Measures survey is also ongoing (http://www.hc-sc. gc.ca/ewh-semt/contaminants/health-measures-sante-eng.php). At present there are several types of biological monitoring guidance values (BGVs), as detailed in Table 1. The German Human Biomonitoring Commission was established in 1992 in order to provide expert advice to the Federal Environment Agency in developing science-based criteria for the application of HBM. The principles and procedures employed by the commission have been previously summarised (Ewers et al., 1996) but in general, a consistent assessment of internal exposure to environmental contaminants is achieved through derivation of two kinds of guidance values, ‘reference’ and ‘HBM’ values (Schulz et al., 2007). Reference values have been defined as ‘the upper margin of the current background exposure of the general population to a given environmental pollutant’ (Angerer et al., 2007). Reference values are not toxicologically-derived biological exposure limits but are a statistical description of the background exposure of a given population at a given time. As environmental conditions and populations are subject to change, reference values need to be updated on a regular basis. HBM values are derived by the Human Biomonitoring Commission following review by an expert panel of toxicological and epidemiological studies. Two levels of HBM values are defined (http://www.umweltbundesamt.de/gesundheit-e/monitor/definitionen.htm); HBM-I values determine the concentration in biological material at which, according to current knowledge, no adverse

Table 1 Examples of human biomonitoring (HBM) guidance values. Acronym

Basis

Organisation

Environmental HBMI HBMII BE Reference value/clinical ranges

RfD/Tox/Epi Tox/Epi/Occ RfD/RfC/MRL/TDI Population ranges

HBM Commission HBM Commission Summitt Toxicology Various

Occupationalª BMGV BAT BEI BLV BALs EKA BLW BAR

Tox/Air/90% range Tox/Air Tox/Air Tox Tox/Air Tox/Air Tox/Epi Ref/Tox

HSE MAK/DFG ACGIH SCOEL WHO DFG DFG DFG

HBMI = human biomonitoring I value; HBMII = human biomonitoring II value; BE = biomonitoring equivalents (Hays et al., 2008); BMGV = biological monitoring guidance value; Reference value = value based on 95th percentile of population range; BAT = biologischer Arbeitsstoff-Toleranzwert (biological exposure value); BEI = biological exposure index; BLV = biological limit value; BALs = biomonitoring action levels; EKA = Expositionsäquivalent für krebserzeugende Arbeitsstoffe (exposure equivalents for carcinogenic substances); BLW = Biologischer Leitwert (biological guidance value); BAR = Biologische Arbeitsstoff-Referenzwerte (Biological Reference Values for Chemical Compounds in the Work Area); RfD = reference dose; RfC = reference concentration; T = tox data; Epi = epidemiology data; Occ = occupational data; Ref = reference range (90% = 90th percentile of occupational survey data). a It should be noted that many countries also publish national HBM guidance values.

health effects are considered to occur. HBM-I values state whether they apply to all groups or whether there are different values for different groups, e.g. children. HBM-II values determine the concentration in biological material above which, according to current knowledge, and judgement of the Commission, and with regard to the substance under consideration there is increased risk of adverse health effects (Schulz et al., 2007). There are also exposure guidance values which are estimates of exposure to chemicals that are considered not to appreciably increase health risk of an exposed population. These values are set by regulatory agencies and authoritative bodies through a risk assessment process. They are designed to protect all groups within the general population, including sensitive subpopulations with chronic exposure. There is a growing body of literature on the derivation of BGVs from toxicological reference values (Hays and Aylward, 2009, 2011; Hays et al., 2008; LaKind et al., 2008). The biomonitoring equivalents (BEs) use toxicokinetic modelling and points of departure based on Acceptable Daily Intake (ADI), tolerable daily intake (TDI) and Provisional Tolerable Weekly Intake (PTWI) data in their derivation. Assessment of human exposure to environmental chemicals forms an integral part of the risk assessment process. Traditionally, exposure is measured or modelled in an environmental medium, and an applied dose for a given cohort estimated. The applied dose is subsequently evaluated against established criteria such as; reference doses (RfDs) or concentrations (RfCs), minimal risk levels (MRLs), tolerable daily intakes (TDIs) or a Unit Cancer Risk (UCR). There are however, some uncertainties involved with estimating environmental exposures, which can lead to both over- and under-estimation of actual exposures (Ewers et al., 1996, 2004; Gosselin et al., 2006). Advances in human biomonitoring in general (biomarker development) and advances in analytical measurements in particular, have allowed the detection and quantitation, at increasingly lower levels, of hundreds of chemicals and their metabolites in biological samples of individuals. However, the ability to interpret these measurements in the context of biological effect has not developed alongside the analytical technologies. As the results from a number of biomonitoring initiatives (described previously) become widely available, there is an increasing need to be able to interpret the results in terms of a public health risk. Such interpretation is essential to achieve efficient communication of any health concern to public health professionals, or to allay concerns of exposed individuals or members of the general population. In addition, it will also impact on decisions regarding risk management and enable prioritisation of resources (Meek et al., 2008). The integration of health and exposure data is the definitive tool in public health. The ultimate goal of any biomonitoring programme should be to collect human health, environmental exposure and biomonitoring data in a fully integrated system. The utility of HBM for public health protection is determined by the ability to interpret the data produced against set criteria; therefore the aim for the future of biomonitoring is to develop a framework for this purpose building on current scientific knowledge and best practice. An accepted framework is therefore urgently needed to help scientists, regulators, and stakeholders to manage the generation and interpretation of HBM data (both for groups and individuals) and to take appropriate action. This appropriate action will include the usual range of risk management options, including risk communication to the public (Sepai et al., 2008). The primary objective of the project reported here was to develop, in consultation with an expert advisory panel, a practical and scientifically-defensible framework for establishing different types of environmental monitoring BGVs for a wide range of environmental contaminants with a range of available data. The framework also aimed to address suggested uses of BGVs and their implications for public health.

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

2. Methods

3. Results

An initial conceptual model of a framework was developed by the project team, which included consideration of the various stages of biomonitoring from detection of a substance to understanding its effects, the different roles of existing guidelines and what could be understood from current biomonitoring data for a given substance. The initial framework was based, in part, on work carried out by a CEFIC funded expert group (ECETOC, 2005) and on the ACC funded work on biomonitoring equivalents (Hays et al., 2007). In particular, criteria for assessing quality of data were as defined by ECETOC (2005). The draft framework was sent out for consultation to an Advisory Group1 comprised of experts who had been invited to participate in the study. Comments received from members of the advisory group were discussed by the project team and the framework amended accordingly. The amended framework was evaluated by expert judgment using 12 chemicals selected to reflect a range of available data (data rich and data poor) and possible levels within the framework. The final list comprised: acrylamide, benzene, cadmium, cypermethrin, di(2-ethylhexyl) phthalate, lead, methyl tert-butyl ether, platinum, polybrominated diphenyl ethers, samarium, toluene diisocyanate, and toluene. Individual framework documents were prepared2 for each of the 12 selected chemicals using expert judgement and weight of evidence, which incorporated consideration of:

3.1. Framework structure and inclusion criteria

 uses, potential routes of exposure and current health concerns;  toxicity data, including biological effect monitoring data, if available (these were not full toxicological reviews but an overview of knowledge);  selection of the most appropriate biomarkers, including information on reference ranges, occupational guidance value, PK or PBPK models, concentration of pure compound or metabolites in blood, urine, environment, inter- and intra-variability;  analytical methods, including information on methods available for measurement of biomarkers (and parent compound if relevant), detection limits, reliability, reference materials, external quality assurance, known confounders;  epidemiology data, including information on any data linking external exposure with health effect, data linking biological monitoring data with health effects;  identification of risk management procedures;  identification of susceptible groups and populations; where a chemical is known to affect a particular subpopulation to a greater extent (either through biology or exposure patterns; IEH, 2002), this should be taken into account; and  communication issues associated with different types of BGVs, different uses, different audiences, including limitations of data (how many, when collected) and of biomarker (half-life, specificity). Framework documents were presented to advisory board members and other invited guests at the workshop3. Comments arising from discussion of the framework and framework documents at the workshop were used to carry out further amendments and the finalised framework structure is presented here.

1 Full framework documents available at: W30 http://www.cranfield.ac.uk/health/ researchareas/environmenthealth/ieh/page21029.html. 2 Full framework documents available at: W30 http://www.cranfield.ac.uk/health/ researchareas/environmenthealth/ieh/page21029.html. 3 Full list of workshop participants and report of workshop available at: W30 http://www.cranfield.ac.uk/health/researchareas/environmenthealth/ieh/ page21029.html.

455

The finalised framework is comprised of four classes or levels of BGV (Fig. 1). Each level of BGV may have several variants depending on the data available and application, but for simplicity it was decided to incorporate only four levels. Moving from left to right across the framework brings an increasing confidence in a health-based guidance value, but also a need for more extensive data; the wavy lines are used to emphasise that the boundaries between levels are not clear-cut. In any framework of this type there will be differences in expert judgement about attribution of a chemical to a particular level. A set of criteria was defined to facilitate assessment of the appropriate level of BGV for a given chemical (further details given in full report (W30) available at http://www.cranfield.ac.uk/health/ researchareas/environmenthealth/ieh/page21029.html). 3.2. Level 1 – biomonitoring data only (not health-based) This level represents the very earliest stage of biomonitoring. Criteria for inclusion in Level 1 include:  availability of an analytical method for the substance (or its metabolites);  detection in a biological fluid (e.g. blood and urine);  absence of toxicity data or evidence of no adverse health effects. 3.3. Level 2 – biomonitoring data plus some concern (not healthbased) This level follows on from Level 1 where there is some interest surrounding possible health effects of the substance. Criteria for inclusion in Level 2 include:  good quality biomonitoring data;  some reason for concern regarding potential health effects based on hazard information;  known health effects following exposure;  some basic metabolism and toxicokinetic data, this could be animal or human;  insufficient data to propose a health-based BGV. 3.4. Level 3 – toxicity data (health-based) Level 3 substances are distinguished from Level 2 through the amount of toxicology data available. Criteria for inclusion in Level 3 of the framework include:  good quality biomonitoring data;  some reason for concern regarding health effects following exposure;  good metabolism and toxicokinetic data, this could be animal or human;  insufficient data to propose a epidemiology-based BGV. 3.5. Level 4 – toxicity and epidemiology data (health-based) Level 4 substances are distinguished from Level 3 by the extent of the data available. Criteria for inclusion in Level 4 of the framework include:  good quality biomonitoring data;  evidence of health effects following exposure;

456

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

Level Level 1

Level Level 2

BM BM data data only only

BM BM data data && interest interest

Population Population range range Either acceptable eClinical clinicalorrange) reference or unknown range

Level Level 3

Not Not HealthHealthbased base Do not exceed, Increasing try to reduce exposure may increase risk

If unknown get population data

BM BM & toxicity toxicit data data

Toxicity & Toxicity & epidemiology probability data data

Health-based Health-based

E.g. BAT, BEI, BE HBM1, ‘Safe’HBM2

Get data sources and toxicology

Level Level 4

‘Safe’ level

Risk-based Risk-based

Safe ‘Safe’level level noadvers adverse outcomes expected expecte

Get epidemiology data to check

No further data needs

Fig. 1. Framework for biomonitoring guidance values. HBM I – the concentration of an environmental toxin in a human biological material (usually blood, serum, plasma, or urine) below which there is, according to the knowledge and judgement of the German MAK commission, no risk for adverse health effects in individuals of the general population; HBM II – the concentration of an environmental toxin in a human biological material (usually blood, serum, plasma, or urine) above which there is – according to the knowledge and judgement of the German MAK commission and with regard to the environmental toxin under consideration – an increased risk for adverse health effects in susceptible individuals of the general population; BAT – biological tolerance value (for occupational exposures) is defined as the maximum permissible quantity of a chemical substance or its metabolites or the maximum permissible deviation from the norm of biological parameters induced by these substances in exposed humans; BEI – biological exposure indices represent conditions under which it is considered (by ACGIH) that nearly all workers may be repeatedly exposed without adverse health effects; BE – biomonitoring equivalents are defined as the biomonitoring levels of specific chemicals in blood, urine or other human biological media or tissues that are consistent with existing exposure guidance values.

 good human metabolism and toxicokinetic data;  good data from epidemiology studies. 3.6. Evaluation of the framework A number of model compounds were used to evaluate the utility of the framework. The criteria met by each chemical and subsequent assigned BGV level, based on expert judgment, is described below and summarised in Table 2. 3.7. Level 1 compounds designated as ‘population range’ BGV 3.7.1. Samarium Measurement of samarium in urine is possible, but there are no published biological monitoring data from either occupational or environmental exposure. There are few data on the toxicity of samarium, however the low acute toxicity of other rare earth elements suggest little immediate concern for environmental exposure. Sources of environmental exposure, toxicokinetic and metabolic data have yet to be defined. There are insufficient epidemiological data to link ill-health for an individual following exposure. 3.8. Level 2 compounds designated as ‘non-health based’ BGV 3.8.1. Methyl tert-butylether Biological monitoring is available based on a number of markers, but for environmental exposures measuring MTBE itself is preferred. Environmental exposures have been determined to a limited extent using both blood and urine samples. There are no environmental or occupational biological monitoring guidance values for MTBE. The toxicology of MTBE is reasonably well studied although there are some studies on carcinogenicity that are currently difficult to interpret for human risk. There have been reports from consumers about acute health symptoms associated with the use of oxygenated fuels such as gasoline containing MTBE. The toxicokinetics and metabolism have been described in both animals

and humans. There are currently insufficient epidemiological data to link ill-health for an individual following exposure. 3.8.2. Platinum Platinum in urine has been included in several HBM studies, and reference ranges are published. Platinum in itself is not very toxic, and the only real platinum-associated health effects have occurred from sensitisation and allergies to specific platinum-salts following occupational exposure. Environmental sources of platinum are primarily from catalytic convertors. Toxicokinetics has been studied in animals with limited human data. There are insufficient epidemiological data to link ill-health for an individual following exposure. 3.8.3. Polybrominated diphenyl ethers HBM data (PBDE in blood) has shown potential wide spread environmental exposure, however reference ranges have yet to be established. Health effects in animals include liver and thyroid toxicity and neurodevelopmental effects. Environmental sources are mainly through its use as a flame retardant. Toxicokinetic and metabolic data are limited. There are insufficient epidemiological data to link ill-health for an individual following exposure. 3.8.4. Toluene diisocyanate HBM for occupational exposure, based on hydrolysable urinary conjugates, has been well reported but there are limited data on environmental exposure. The occupational biological guidance values set are not health based. TDI has been classified as a class 2B carcinogen (IARC, 1999) and it is known to be a respiratory irritant and sensitiser. There are some plausible routes of exposure to the public through leaks and emissions near and around industrial plants manufacturing TDI as well as from polyurethane production. In addition there are some concerns around indoor air quality and low level emission of reactive isocyanates from consumer products. Toxicokinetics have been studied in animals with limited human data. There are insufficient epidemiological data to link ill-health for an individual following exposure.

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

457

4 4 3

U

3

U

2 2 2

3.9.3. Di(2-ethylhexyl)phthalate Exposure of the general population to DEHP is omnipresent (due to its use as a plasticiser) and well documented by human biomonitoring data. The oxidised metabolites, 5OH-MEHP, 5oxoMEHP and 5cx-MEPP are the biomarkers of choice. The primary health-effect in animals is reproductive toxicity. Both human and animal metabolism and toxicokinetics of DEHP are well described. There are recent epidemiological studies that indicate subtle effects of DEHP even at environmental exposures; however, it was judged that findings from these were not sufficient to propose an epidemiology-based guidance value.

U – data available;  – data not available.

3 1

2

   



3.9.2. Cypermethrin Biological monitoring based on urinary metabolites is well established and reported at environmental levels, with an external quality assurance scheme available. The German Human Biomonitoring Commission has established a reference range of urinary metabolites for the German population and there are also data from the NHANES studies in the US and other countries. Adverse health effects have been reported in humans after acute exposure, with neurological effects apparent in animal studies. Environmental exposure is generally through contaminated food and use of pesticide products. The toxicology of cypermethrin, its toxicokinetics and metabolism have been well described in both animals and humans and a PBPK model has been developed. There are insufficient epidemiological data to predict the probability of ill-health for an individual following exposure.

3

  U   

Epidemiology Link between exposure and human health effect Link between BM data and human health effect Level





U  U U U  U U U Analytical Detection limits Ref. materials/QA scheme Confounders

U  

U  

U   U U  U  U    Biomarkers Ref. ranges Occupational guidance values PK/PBPK models

U  U

  U U   U  U U  U    Human toxicity Toxicokinetics Dose–response Main metabolites

3.9.1. Acrylamide Numerous studies show estimation of internal exposure can be reliably carried out through measurement of the Hb-adduct of acrylamide (AAVal) which reflects the acrylamide dose taken up in the previous 4 months. Reference ranges have been established for AAVal for adults and children. Two mercapturic acids of acrylamide are the main metabolites, which are excreted with urine and represent exposure of the last 2 days. Acrylamide is a suspected carcinogen in humans, with the main sources of acrylamide exposure being food and tobacco smoke; acrylamide therefore poses a health risk. Metabolism and toxicokinetics of acrylamide have been defined in both humans and animals. There are insufficient epidemiological data to predict the probability of ill-health for an individual following exposure.

4

U U

U U U U  U U  U

U

U

U

U U U

U

U U U U U U U U U

U  U

U U U U U U U U  U U 

U  

U U U

U U 

U U U U U U U U U U U U

U  U

U U U

U U U

U U U U U U U  U U U  U  U U U  Exposure General population Susceptible groups

U 

U 

Decabrominateddiphenyl ether Platinum Methyl tertbutylether Cypermethrin Samarium Framework element

Table 2 Summary of data availability for model compounds.

U U

Benzene Acrylamide Toluene diiosocyanate

Di(2ethylhexyl)phthalate

Toluene

Cadmium

Lead

3.9. Level 3 compounds designated as ‘toxicity/health-based’ BGV

3.9.4. Toluene Measurement of toluene in blood or urine is specific for toluene exposure. A biomonitoring equivalent (BE) value for environmental exposure to toluene based on levels in blood has been derived. Population data show that although environmental exposure to toluene is widespread, the levels found are below the BE value. Toluene is not carcinogenic and the lead health effects are CNS depression and foetal toxicity. The metabolism and toxicokinetics of toluene are well described. There are insufficient data to describe the long-term effects of low level environmental exposure to toluene or to propose a probability-based guidance value. 3.10. Level 4 compounds designated as ‘toxicity and risk-based’ BGV 3.10.1. Benzene Extensive HBM data on occupational and environmental exposure to benzene are available. A sensitive and specific urinary biomarker (S-phenylmercapturic acid) is available and has been widely used in environmental surveys. External quality assurance is available. Known health effects in humans following chronic exposure are haematoxicity, genotoxicity and carcinogenicity.

458

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

The main sources of benzene exposure in ambient air include, cigarette smoke, petrol (combustion and evaporation), petrochemical industries and other combustion processes. Metabolism and toxicokinetics of benzene have been reported in both humans and animals. There are epidemiological data to predict the probability of ill-health for an individual following exposure. 3.10.2. Cadmium Urinary cadmium concentrations are an excellent biomarker to describe exposure to cadmium and can directly be related to associated health effects. Reference values and health based HBM-values for the general population have been established. The kidney has been identified as a key organ in cadmium-associated health effects and cadmium has been classified as a lung carcinogen. Environmental exposure is widespread from smoking, contaminated food, water and air. The primary pathways for uptake of cadmium, toxico- and pharmacokinetic properties, and associated health effects are well understood and have been validated under different exposure scenarios. Epidemiological studies have confirmed the in vivo and in vitro findings on the cadmium exposure-dose–response continuum in the general population. 3.10.3. Lead Lead is probably the best-studied environmental pollutant with respect to the use of HBM values in hazard identification, hazard/ risk assessment, risk reduction strategies and risk management evaluation. Measurement of blood lead has been used to assess environmental exposure for over 40 years. Health effects include neurological, reproductive and haematological toxicity. Lead has also been classified as a Group 2A carcinogen (IARC, 2006). The most important sources of lead in the environment (tetra-ethyl lead in gasoline and lead-containing paint) are being banned across the world. The epidemiological consequences of population-scale exposure to lead through different sources are widely understood and supported by a substantial amount of scientific evidence. Risk reduction and management strategies are available, have proven their efficacy, and are globally applied. 4. Discussion The ability to measure substances in biological fluids often exceeds our understanding of their relevance. The primary objective of the study reported here was to develop a practical and scientifically-defensible framework for managing the interpretation of HBM data. This framework, initially developed by the research team based on their experiences and the existing literature, was modified following critical comments from an expert advisory board, evaluated using a number of substances (data rich and data poor) and finally, fine-tuned following presentations and discussions at an expert workshop. A number of environmentally and occupationally-derived biological monitoring guidance values are currently utilised for biomonitoring. These BGVs are most well developed in the occupational context where they have been used for many decades and where exposures usually have been many orders of magnitude above exposures environmentally encountered. Many of the occupational BGVs are health-based but others, where exposure-effect information is not available, may be more pragmatically derived and established on good practice and used as a risk management tool in the workplace. To some extent, there are parallels to environmentally-determined BGVs. However, to acknowledge that not all types of BGVs are derived or applied in the same way, the framework developed within this study is comprised of four levels (Fig. 1). The level of BGV assigned to a given compound is dependent on expert judgement and reflects the type and quality of

available HBM data and potential health/toxicological effects, and its relevance (or otherwise) to health. Assignment to a particular level could be a source of discussion between experts at the boundaries between levels. Although moving through the framework from Levels 1 to Level 4 brings an increased confidence and moves towards a health-based guidance value, in practice, assigning a substance to Levels 3 and 4 requires extensive data. Such data is only achieved through studies or surveys which may take many years to achieve (equivalent to risk assessments and epidemiology studies) and are usually only carried out for substances with widespread exposure and significant concern regarding health effects. The majority of substances are likely to fall into Levels 1 and 2 of the framework, as HBM data is most likely to be comprised of a simple range of observed values. These two levels of the framework allow development of BGVs based on data representing the distribution of a substance in the general population and exposure-based BGVs (equivalent to NHANES or German Reference Values) to be developed. BGVs at Levels 1 and 2 can provide valuable data on ‘background’ levels in a given population, and aid identification of high exposure groups. Each level of BGV within the framework is associated with a possible action that may be required to move a substance to the next level(s). It is considered that in the main, the need to undertake recommended actions would be as a result of justified interest or concern over any potential risk associated with exposure to that substance. However, for some substances it is possible that no further action or movement between levels would be necessary. In such cases, it needs to be made clear that a substance placed in Levels 1 and 2 may not be in such a category simply because of lack of information, and that there is not an automatic need to gather more HBM data. It might well be the case that the substance is there because it is toxicologically of very little or no concern, human exposures are likely to be low in known and predicted circumstances and thus Level 1 may the natural and permanent home for that substance. The framework presented here is to be seen as part of an iterative process for the use of HBM to monitor environmental natural and anthropometric substances in the general populations. To date, only a very limited number of countries (Germany and the US) seem to have developed systematic national HBM programmes for environmental exposure which assist in underpinning public health policy. However, this activity is likely to grow and a framework, such as the one outlined here, may well change over time. This is to be expected as the framework is intended to assist in the interpretation and communication of HBM information. With usage, more experience and more HBM information from surveys and studies, it may well evolve to better meet the needs of those who need to interpret and explain to various stakeholders what the information might or might not mean in terms of public health. The framework is built on previous work such as the Guidance for the Interpretation of Biomonitoring Data (ECETOC, 2005) which identified elements of knowledge (analytical integrity, toxicokinetics, health effects and weight of evidence) in order to set the ‘‘boundary of interpretability’’. As in our framework, ECETOC defined four levels – exposure trends (biomarker data), exposure characterisation, health impact assessment and risk assessment and standard setting. Although the ECETOC report noted ethical and communication considerations, these were not part of the structure outlined whereas our framework includes communication and also risk management measures where available and appropriate, allowing action to be taken in cases where exposures are deemed too high (particularly for Level 4 compounds where the determination of over-exposure and the associated risks are known). ECETOC recommended that their framework should be

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

further validated and as ours reflects similar categories and a similar approach, it could be argued that we have done this as well as expanded the framework. ECETOC also concluded that there was a need for more health-based biomonitoring guidance values for the general population; the increasing production of biomonitoring equivalents (Hays et al., 2007, 2008; Hays and Aylward, 2009, 2011; LaKind et al. 2008); has developed this extensively and there are now general population guidance values for over 80 chemicals (Angerer et al., 2011). This expansion of guidance values based on biomonitoring equivalents or similar approaches (e.g. HBM I and II) increases the number of guidance values within Level 3, meaning that many biomarker results can be interpreted on a group basis with some certainty as to whether ‘‘over-exposure’’ has occurred however the number of chemicals in Level 4, where there is good human epidemiology data, is likely to remain low for the foreseeable future. There is a growing trend for chemicals with some ‘‘concern’’ to be restricted in use (e.g. phthalates, bisphenol A) before such Level 4 data is available. The approach taken and reported here complements that of Sobus et al. (2011). In a recent paper the authors presented a five-tiered framework designed to improve the interpretation and use of biomonitoring research to support risk assessments. Each tier within the framework describes different uses for biomarkers with different degrees of complexity. Tier 1 is based on biomarker data alone, the same as for our Level 1 compounds. In the framework described by Sobus et al., additional measurements and estimated values (from modelling) are added in subsequent tiers, for example, correlations with environmental exposures in Tier 2 up to biologically-relevant dose estimates in Tier 4. As with our system, moving through the Tiers requires increasing knowledge and data with initial levels involving exposure surveillance and assessment, progressing to risk assessments at the higher levels. However, the framework presented by Sobus et al., is focussed on determining and interpreting dose estimates of exposure from measured biomarker levels whereas ours is more directed at the level of confidence of all data (biomarker, toxicology, epidemiology) used to determine a particular biological monitoring guidance value. At the current time, biomonitoring is seen as a valuable tool that is used to provide information on exposure of the general population to a variety of environmental chemicals. However, there remain several scientific and technical limitations of biomonitoring that need to be addressed. Of these, integration of health and exposure data is critical, with collection of human health, environmental exposure and biomonitoring data in an integrated system being the ultimate goal of any biomonitoring programme. The utility of HBM for public health protection is determined by the ability to interpret the data produced against set criteria, also a key limitation at present. However, the absence of ‘all the answers’ should not inhibit the search for them. Rather, an accepted framework is needed to put the available information in perspective, manage the expectations of its use and inform the need for additional data. Ideally, such a framework would be based on current scientific knowledge and best practice across Europe, the US and elsewhere, to help scientists, regulators, and stakeholders to design appropriate HBM studies, interpret HBM data (both for groups and individuals) and to take appropriate action. This appropriate action will comprise a range of risk management options, including risk communication to the public. For Level 1 substances communication and interpretation will be limited due to insufficient data to determine risk, unless data falls within clinical reference ranges (this will only apply to a limited number of substances). For Level 4 substances, there will be substantial information to inform risk communication and risk management options; however implementation of risk management measures becomes a political and societal debate.

459

There are many issues with regard to communication that are not unique to HBM but relate to the conduct of all types of health-related research in human subjects. The issues of greatest importance range from ‘transparency’ and openness with all stakeholders to risk communication. As with all such communication, it should not be treated as a bolt-on extra to be considered only at the end of the research or survey, but at all stages of the process and thus seen as an integral part of the overall activity. On the other hand, there are issues specific to human biomonitoring and these are related to how biomarkers of exposure and effect are interpreted in the context of individual health risk and population health risk management. The rapid advances in analytical technologies have resulted in a vast array of exposure and effect data sets of varying size and quality. However, not only has our ability to interpret these data fallen behind but so has our ability to communicate what these biomarker levels in human tissue may mean in terms of health. The framework described here may be used as a tool to aid clear and transparent communication to a wide range of stakeholders. These stakeholders include the public at large, the study population, the scientific community as well as politicians, the media and other interested parties. Each of these audiences has a different goal, different needs and different perception of risk. As an example the detection of a substance in a biological fluid for the first time (Level 1) is not interpretable in terms of health for an individual, and this limitation should be communicated in data gathering studies. In contrast, measurement of the concentration of a substance in Level 4 is interpretable in terms of risk of ill health to the individual.

5. Conclusions The framework presented here is a practical and scientificallydefensible framework for managing the interpretation of HBM data, based on available evidence. Several key features are included:  a set of criteria for acceptability of data, consideration of applicability and any associated action;  all substances with HBM data can be included, and an appropriate level of BGV identified;  guidance values used to derive the level of BGV are considered in context of their application, thereby preventing misinterpretation;  clear consideration is given to the interpretation of BGVs, with limitations clearly defined; and  communication of BGVs to a wide range of stakeholders including, general public, the study population, the scientific community, politicians, media and other interested parties is considered for substances at the individual level. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgements We gratefully acknowledge the support of CEFIC in funding this study (LRI-HBM1-UCRA-0712) and Dr. Peter Boogaard for acting as Project Officer. Thanks to Dr. Silvia Fustinoni for her input on the chemical Methyl tert-butylether, the advisory board for their constructive comments to the project team during framework development and to workshop participants for their input to the final framework.

460

R. Bevan et al. / Regulatory Toxicology and Pharmacology 63 (2012) 453–460

References Angerer, J., Aylward, L.L., Hays, S.M., Heinzow, B., Wilhelm, M., 2011. Human biomonitoring assessment values: approaches and data requirements. International Journal of Hygiene and Environmental Health 214 (5), 348–360. Angerer, J., Ewers, U., Wilhelm, M., 2007. Human biomonitoring: state of the art. International Journal of Hygiene and Environmental Health 210 (3–4), 201–228. ECETOC, 2005. Guidance for the Interpretation of Biomonitoring Data. Document No. 44. European Centre for Ecotoxicology and Toxicology of Chemicals, Bruseels, Belgium. 1–71pp. Ewers, U., Boening, D., Albrecht, J., et al., 2004. Gefährdungsabschätzung von Bodenverunreinigungen in Einem Belasteten Wohngebiet: Bedeutung und Funktion des Human-Biomonitoring – Ein Fallbeispiel (Risk assessment of soil contamination in a residential area: the importance and role of human biological monitoring – a case report). Gesundheitswesen 66 (8–9), 536–544 (cited in Hays et al. (2007), German). Ewers, U., Wittsiepe, J., Schrey, P., et al., 1996. Levels of PCDD/PCDF in blood fat as indices of the PCDD/PCDF body burden in humans. Toxicology Letters 88 (1–3), 327–334 (cited in Hays et al. (2007)). Gosselin, N.H., Brunet, R.C., Carrier, G., et al., 2006. Reconstruction of methylmercury intakes in indigenous populations from biomarker data. Journal of Exposure Science & Environmental Epidemiology 16 (1), 19–29. Hays, S.M., Aylward, L.L., 2011. Interpreting human biomonitoring data in a public health risk context using biomonitoring equivalents. International Journal of Hygiene and Environmental Health (Epub ahead of print). Hays, S.M., Aylward, L.L., 2009. Using biomonitoring equivalents to interpret human biomonitoring data in a public health risk context. Journal of Applied Toxicology 29 (4), 275–288. Hays, S.M., Aylward, L.L., LaKind, J.S., et al., 2008. Guidelines for the derivation of biomonitoring equivalents: report from the biomonitoring equivalents Expert Workshop. Regulatory Toxicology and Pharmacology 51 (3 Suppl. 1), S4–S15. Hays, S.M., Becker, R.A., Leung, H.W., et al., 2007. Biomonitoring equivalents: a screening approach for interpreting biomonitoring results from a public health risk perspective. Regulatory Toxicology and Pharmacology 47 (1), 96–109. Heinrich-Ramm, R., Jakubowski, M., Heinzow, B., 2000. Biological monitoring for exposure of volatile organic compounds (VOC). Pure Appl Chem 72, 385–436.

IARC, 2006. Inorganic and Organic Lead Compounds. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 87. International Agency for Research on Cancer, Lyon, France. IARC, 1999. Re-Evaluation of some Organic Chemcials, Hydrazine and Hydrogen Peroxide. Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 71, IARC, Lyon, France. Indulski, J., Jakubowski, M., Merecz, D., 1999. Present environmental lead exposure in Polish children and its possible influence on the CNS functions. European Journal of Oncology 4, 631–636 (cited in ECETOC, 2005). Institute for Environment and Health (IEH), 2002. Variability and Susceptibility in Human Response to Occupational Exposure in the UK (Report 13). MRC Institute for Environment and Health, Leicester, UK. Jakubowski, M., Trzcinka-Ochocka, M., 2005. Biological monitoring of exposure: trends and key developments. Journal of Occupational Health 47 (1), 22–48. LaKind, J.S., Aylward, L.L., Brunk, C., et al., 2008. Guidelines for the communication of biomonitoring equivalents: report from the Biomonitoring Equivalents Expert Workshop. Regulatory Toxicology and Pharmacology 51 (3 Suppl. 1), S16–S26. Meek, M.E., Sonawane, B., Becker, R.A., 2008. Foreword: biomonitoring equivalents special issue. Regulatory Toxicology and Pharmacology 51 (3 Suppl.), S3. Schulz, C., Angerer, J., Ewers, U., et al., 2007. The German Human Biomonitoring Commission. International Journal of Hygiene and Environmental Health 210 (3–4), 373–382. Sepai, O., Collier, C., Van Tongelen, B., et al., 2008. Human biomonitoring data interpretation and ethics; obstacles or surmountable challenges? Environmental Health 7 (Suppl. 1), S13. Sobus, J.R., Tan, Y.-M., Pleil, J.D., et al., 2011. A biomonitoring framework to support exposure and risk assessments. Science of the Total Environment 409, 4875– 4884. US EPA – United States Environmental Protection Agency, 2006. Reregistration Eligibility Decision for Cypermethrin. Available at: . Wilson, N.K., Chuang, J.C., Lachan, R., et al., 2004. Design and sampling methodology for a large Study of Preschool Children’s Aggregate Exposures to Persistent Organic Pollutants in their Everyday Environments. Journal of Exposure Analysis and Environmental Epidemiology 14 (3), 260–274.