SummaryAn expert workshop on effects of combined exposure to chemicals, with specialemphasis on chemicals with endocrine activity was held under the auspices of the DanishMinistry of the Environment. The aim of the workshop was to examine existing scientificknowledge on combination effects of endocrine disrupters, with a focus on regulatoryaspects. The workshop participants considered the state of the science of mixtures riskassessment for endocrine disrupters, and discussed the feasibility of approaches tocumulative risk assessment.A consensus about a number of important issues could be formulated, and this included aseries of recommendations:Cumulative risk assessment (CRA) for endocrine disrupters was seen as both necessaryand feasible. The predominant chemical-by-chemical approach in risk assessment wasregarded as insufficiently protective against the possibility of mixture effects/ effects ofcombined exposure.The application of dose (or concentration) addition as an assessment method wasrecommended as a default, until evidence as to the suitability of alternative assessmentconcepts emerges.A pre-occupation with mechanisms or modes of action as the starting point for thegrouping of endocrine disrupters into classes to be subjected to mixtures risk assessmentwas seen as not practical and scientifically hard to justify. Instead, grouping criteriashould focus on common health related effects and the likelihood of co-exposures.The full potential of CRA for endocrine disrupters cannot be reached without filling anumber of data gaps, most importantly in the area of mixtures exposure assessment.An enhancement of the legal framework in Europe with a view to mandating CRA shouldbe given serious consideration.

Expert workshop on combination effects of chemicals, Hornbæk, DenmarkAbbreviationsADIAhRBBPCACERCLACMGCRADADBPDEHPDEPAERFQPAGHSHIIAMOENOECNOELNOAELNRCPBDEPCDD/FPODIREACHRfDTCDDTDITEFTEQUFUVBCAcceptable daily intakeAryl hydrocarbon receptorBenzyl butyl phthalateConcentration additionComprehensive Environmental Response Compensation and Liability ActCommon mechanism groupCumulative risk assessmentDose additionDibutyl phthalateDiethyl hexyl phthalateDanish Environmental Protection AgencyEstrogen receptorFood Quality Protection ActGlobal Harmonisation SystemHazard indexIndependent actionMargin of exposureNo observed effect concentrationNo observed effect levelNo observed adverse effect levelNational Research CouncilPolybrominated diphenyl etherPolychlorinated dibenzo-p-dioxin/furanPoint of departure indexRegistration Evaluation and Authorisation of ChemicalsReference doseTetra chloro dibenzo-p-dioxinTolerable daily intakeTCDD equivalency factorTCDD equivalentUncertainty factorUnknown or variable composition complex reaction products or biologicalmaterials

1. Terms of reference and workshop aimsThe Danish Environment Minister authorized the Danish Environmental ProtectionAgency (DEPA) to host an expert workshop on combination effects of chemicals, withspecial emphasis on endocrine disrupters. This workshop took place on 28 – 30 January2009 in Hornbæk, Denmark.The aim of the workshop was to examine existing scientific knowledge on combinationeffects of endocrine disrupters, with a focus on regulatory aspects. The followingquestions were to be addressed:•••••What is the state-of-the-science on combination effects at present – for chemicalsin general and specifically for endocrine disrupters?Which problems can be identified on the basis of the existing knowledge – inrelation to health and in relation to the environment?What are the challenges the regulatory authorities have to face?How can these challenges be met and the existing knowledge be taken intoaccount within the existing regulation?What are suggestions for actions with a focus on regulatory aspects on a global, aregional (EU) and national (DK) level?

2. The workshop programme, resource materialsTo realize the workshop aims, five different sessions were set up.Session 1,“Mixtures risk assessment – is it necessary?” was intended as a first steptowards defining the issues of the workshop. A second goal was to review theexperimental evidence for mixture effects when chemicals are combined at low doses,close to levels that are “points of departure” for risk assessment and regulation (e.g.benchmark doses or NOAELs).The plan forSession 2,“A basis for combined risk assessment – case study: phthalatesand other anti-androgens” was to summarize the experimental evidence for combinationeffects of antiandrogens, to review criteria for grouping these substances for purposes ofmixtures risk assessment and to gain an overview of risk assessment methods formixtures.Session 3,“The basis of combined risk assessment for other classes of endocrinedisrupters and other chemicals” aimed to consider topics for mixture risk assessmentrelevant to other endocrine disrupting chemicals, such as: What are effect outcomes ormechanisms on which mixtures risk assessment should be based? What is the evidencefor combination effects?Session 4,“From mixtures risk assessment to regulation” was set up for a more generaltreatment of the mixtures risk assessment, relevant to other groups of chemicals, bysummarizing approaches to mixtures regulation, also in ecotoxicology, including ananalysis of uncertainty factors and their suitability for dealing with mixture effects.

Expert workshop on combination effects of chemicals, Hornbæk, DenmarkSessions 1 – 4 consisted of a series of formal talks, followed by discussions. The talkswere based on resource material which was distributed in advance to all participants.Finally,Session 5“Looking forward – what can/should be done?” was conducted in theform of a structured discussion among workshop participants, with the aim of drawing uprecommendations for risk assessment, regulation and research.The workshop programme together with bibliographic references for the resourcematerial, and the list of participants can be found in the appendix. Since most formal talksduring the workshop were based on published scientific articles, their content isaccessible through the resource list. For this reason, the workshop talks will not besummarized in chronological order in this report. Rather, a structured digest of thepresentations, discussions and recommendations of the workshop will be given.

3. State of the science on combination effects of endocrinedisruptersOver the past decade, mixture toxicology has undergone a remarkable and productivedevelopment. While earlier experimental studies focused mainly on mixtures composedof only two chemicals, the planning, conduct and assessment of multi-componentmixtures with up to 50 chemicals is now state of the art. This has extended from in vitroassays to in vivo studies, although scientific data about in vivo combination effects areless prevalent than in vitro studies.Most mixture studies with endocrine disrupters published in the peer-reviewed literaturehave been conducted with the aim of explaining the joint action of selected purecompounds in terms of their individual effects (component-based approach).3.1 Definitions and termsIt is noted that the terms “mixture effects” and “effects of combined exposure” (to morechemicals) are used without discrimination here and that the term “mixture” thus has abroader meaning in this context than when used in chemicals legislation includingguidance (e.g. REACH and GHS). The field of mixture toxicology is notorious for its useof poorly defined terms. Depending on context, there are many synonyms, and someterms are uncritically used with entirely different meanings. For this reason, workshopparticipants agreed on tentative definitions for a number of frequently used terms:Mixture:A mixture is a combination of several chemicals with which organisms comeinto contact, either simultaneously, or sequentially. A binary mixture is a combination oftwo agents. The term “complex mixture” is used to denote a mixture of unknowncomposition, isolated from environmental media or other sources. “Complex mixture” issometimes used to describe combinations composed of three or more chemicals, but forthe purposes of this review, the term “multi-component mixture” is preferred.Mixture effect, combination effect, joint effects:The response of a biological system toseveral chemicals, either after simultaneous or sequential exposure. The terms are usedsynonymously.

Expert workshop on combination effects of chemicals, Hornbæk, DenmarkAdditivity:In the context of mixture toxicology, additivity cannot be equated with“additivity” in the mathematical sense. It refers to a situation, termed “non-interaction”(and often used synonymously with ”additivity”), where the toxicity of a mixtureresembles the effects expected to occur when all mixture components act withoutdiminishing or enhancing each others effects. Additivity expectations for mixtures can bederived from the concepts of dose (or concentration) addition and independent action (see3.2.1 and 3.2.2). In certain situations, valid expectations for additive combination effectscan also be calculated by building the arithmetic sum of the individual effects of allmixture components (“effect summation”).Synergism, antagonism:When an observed combination effect is larger or smaller thanexpected according to an additivity assumption (based either on dose addition orindependent action), there is synergism or antagonism, respectively.Mechanism of action:Molecular sequence of events that produce a specific biologicalresponse.Mode of action:A sequence of key cellular and biochemical events with measurableparameters that result in a toxic effect. Mode of action considerations are used to decidewhether an effect observed after administration of a chemical in animals has relevancefor humans. Mode of action is not intended to build a comprehensive model of achemical’s actions. It is often confused with mechanism of action, or used in overlappingways. Mode of action can include mechanisms of action, but is considered to be broader.Cumulative risk assessment (CRA), mixtures risk assessment:The terms are usedsynonymously. They denote risk assessment approaches that consider the impact ofmultiple chemical exposures, from multiple sources, routes and pathways, over multipletime frames. It is worth noting that the European use of the term “cumulative riskassessment” encompasses multiple sources, routes and pathways, but restrictsconsiderations to one chemical, not multiple chemicals. For the purposes of this report,the European use of the term is ignored. Toxicity assessments of multi-constituentsubstances (e.g. technical solvents) or UVBC (unknown or variable composition,complex reaction products or biological materials) also do generally not fall undermixtures risk assessments of the kind discussed during the workshop. The reason is thatmulti-constituent substances and UVBCs often are treated in the same way as a singlechemical entity would be dealt with; no attempts are made to explain mixture effects interms of the activity of the constituents.There are various approaches to chemicals risk assessment (Suter and Cormier 2008), andthese also impact on CRA. First, risk assessment can be carried out in order to providetrigger values for regulatory action to protect humans or wild life from harm(“protective” risk assessment). In this case, a bias towards conservatism and worst caseassumptions is essential. Second, there is risk assessment aimed at quantifying themagnitude of impact resulting from certain exposures to chemicals. Such approaches(“quantitative” risk assessment) need to be as accurate as possible in their risk estimatesand tend to utilize probabilistic methods. This report is mainly concerned with protectiverisk assessment, and less so with quantitative risk assessment.

Expert workshop on combination effects of chemicals, Hornbæk, Denmark3.2 Prediction and assessment of mixture effectsWhen several chemicals occur together in a mixture, they may influence each otherseffects by enhancing or diminishing their action. In mixture toxicology, such situationsare described as toxic interactions. More frequently however, chemicals act togetherwithout influencing each others actions. In such cases, it is possible to anticipatequantitatively the effects of a mixture from knowledge about the effects of its individualcomponents. This phenomenon is called non-interaction or additivity. Two concepts areavailable for the formulation of the null hypothesis of additivity:dose (or concentration)additionandindependent action.These concepts are based on two entirely different ideas about how the joint action ofchemicals can be perceived.3.2.1 Dose additionDose addition (DA) is based on the idea that all components in the mixture behave as ifthey are simple dilutions of one another, which is often taken to mean that the conceptdescribes the joint action of compounds with an identical mechanism of action. Whenthese chemicals interact with an identical, well-defined molecular target, it is thought thatone chemical can be replaced totally or in part by an equal fraction of an equi-effectiveconcentration (e.g. an EC50) of another, without changing the overall combined effect.A widely used application of this approach is the “toxic equivalence factor” (TEF)concept for the assessment of mixtures of polychlorinated dioxins and furans (PCDD/F)(van den Berg et al. 1998). Here, doses of specific PCDD/F isomers are all expressed interms of the dose of a reference chemical, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD),needed to induce the same effect (“equivalent” or “equi-effective” dose). The assessmentof the resulting combined effect is obtained simply by adding up all equivalent TCDDdoses. The application of TEF only holds when the underlying dose-effect relationshipsare linear. If this pre-condition is violated, TEFs vary with the effect level that isconsidered for analysis.DA implies that every toxicant in the mixtures contributes in proportion to its toxic unit(i.e. its concentration and individual potency) to the mixture toxicity. Whether theindividual doses are also effective alone does not matter. Thus, combination effectsshould also result from toxicants at or below effect thresholds, provided sufficiently largenumbers of components sum up to a sufficiently high total dose. In view of the exposuresituation in many environmental compartments, the verification or falsification of thisconclusion has been a major topic in recent mixture toxicity studies (see below). Anoverview of mixture studies that focused on this issue is given by Kortenkamp and co-workers (Kortenkamp et al. 2007).3.2.2 Independent action (response addition, effect multiplication)Independent action (IA) conceptualises mixture effects in a different way. It assumes thatthe joint effect of a combination of agents can be calculated from the responses ofindividual mixture components by adopting the statistical concept of independent events.The resulting combined effect can be calculated from the effects caused by the individualmixture components by following the statistical concept of independent random events(Bliss, 1939).

Expert workshop on combination effects of chemicals, Hornbæk, DenmarkAs IA uses the individual effects of the mixture components to calculate the expectedmixture effect, this concept implies that agents present at doses below their individualeffect thresholds (i.e. at zero effect levels) will not contribute to the joint effect of themixture. Hence if this condition is fulfilled for all components there will be nocombination effect. This central tenet of IA is commonly taken to mean that exposedsubjects are protected from mixture effects as long as the doses of all agents in thecombination do not exceed their no-observed-effect-levels or –concentrations (NOEL orNOECs) (see below).3.2.3 Choosing between dose addition and independent action for the purpose ofassessment and predictionA question of fundamental importance to risk assessment and regulation is which of thetwo concepts, DA or IA, should be exclusively chosen for the interpretation of empiricaldata, or for anticipating mixture effects of untested combinations. As a way of resolvingthe issue, DA and IA have been allied to broad mechanisms of combination toxicity, withDA thought to be applicable to mixtures composed of chemicals with a similar mode ofaction, with the corresponding mechanistic model of “simple similar action”, and IA forchemicals with diverse modes of action, and the mechanistic model of “independent jointaction”.The issue of distinguishing between these mechanistic models becomes especiallyimportant, when DA and IA predict different mixture toxicities. In such cases it isimportant to realize that the prediction differences or similarties stem from themathematical features that form the basis of DA and IA (Drescher and Boedeker 1995).Prediction differences are not driven by the biology or toxicology of combinations ofchemicals with similar or diverse mode of actions.Dose addition is thought to be applicable to mixtures composed of chemicals that actthrough a similar or common mode of action (US EPA 1986, 1999, 2000). Although theoriginal paper by Loewe and Muischneck (1926) contains little that roots dose addition inmechanistic considerations, the idea of similar action probably derives from the“dilution” principle which forms the basis of this concept. Because chemicals are viewedas dilutions of each other, it is implicitly assumed that they must act via common orsimilar mechanisms.Conversely, IA is widely held to be appropriate for mixtures of agents with diverse or“dissimilar” modes of action. Although rarely stated explicitly, this presumably stemsfrom the stochastic principles that underpin this concept. The idea that chemicals actindependently is equated with the notion of action through different mechanisms. Byactivating differing effector chains, so goes the argument, every component of a mixtureof dissimilarly acting chemicals provokes effects independent of all other agents thatmight also be present, and this feature appears to lend itself to statistical concepts ofindependent events. However, theoretically, the stochastic principles of IA are also validwhen one and the same agent is administered sequentially. This can be illustrated byusing cytotoxicity as an example. Because cells cannot die twice, the probabilisticprinciple of IA applies, even though the precise mechanisms that underlie the cytotoxicaction of the chemical are identical in sequential administration. In the case ofsimultaneous administration of many chemicals however, the principle of independent

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkevents only applies when the additional assumption is made that all mixture componentsact strictly independently, through different mechanisms.The practical relevance of IA for the assessment of mixture effects has been called intoquestion on the basis of considerations of biological organisation. The principle of strictlyindependent events may rarely apply due to converging signalling pathways and inter-linked subsystems. For these reasons, DA is seen as more broadly applicable, and hasbeen termed the “general solution” for mixture toxicity assessment (Berenbaum 1985).However, the few studies that were specifically designed for a comparative evaluation ofboth concepts for mixtures composed of strictly dissimilarly acting substances,demonstrated that IA provides a better prediction of the observed mixture toxicities(Backhaus et al. 2000; Faust et al. 2003). These observations argue against DA as the“general solution” for mixture assessments.It appears that theoretical considerations are not decisive in answering the question ofchoice between DA and IA as assessment concepts for endocrine disrupter mixtures. Toresolve the issue, it is therefore necessary to consider the empirical evidence.3.3 Dose addition or independent action? - Experimental evidence with mixtures ofendocrine disruptersThe study of mixtures composed of chemically pure endocrine disrupters, in laboratorysettings, has yielded a considerable body of evidence showing that concentration (dose)addition provides a sound approximation of experimentally observed additivecombination effects (see the review by Kortenkamp 2007). However, due to apredilection of researchers to combine endocrine disrupters of the same type (e.g.estrogenic, antiandrogenic or thyroid-disrupting chemicals), in many of the publishedstudies IA could not have been expected to produce valid additivity expectations.Even so, there are recent indications that DA gives better approximations of combinationeffects of endocrine disrupters with diverse modes of action. For example, Rider et al.(2008) conducted mixture experiments with the three phthalates BBP, DBP, and DEHP incombination with the antiandrogens vinclozolin, procymidone, linuron, and prochloraz.Its components have a variety of antiandrogenic modes of action. Vinclozolin andprocymidone are AR antagonists, and linuron and prochloraz exhibit a mixed mechanismof action: inhibiting steroid synthesis and blocking the steroid receptor. DA gavepredictions of combined effects of the mixed-mode antiandrogens that agreed better withthe observed responses than did the expectations derived from IA.Mixtures of thyroid disrupting chemicals with diverse modes of action also showedcombination effects that were approximated better by DA, not IA (Kevin Crofton,workshop presentation of unpublished data).No case has yet been identified, where IA yielded predictions of endocrine disruptercombination effects larger than those derived from DA, and at the same time were inagreement with experimental data. Taken together, the determinants of additive jointaction of endocrine disrupters are fairly well established, and it appears that DA providesgood approximation of combination effects. Therefore, until evidence to the contrary

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkemerges, DA can be adopted as the default concept for the assessment and prediction ofendocrine disrupter mixture effects.Factors that might lead to deviations from expected additive effects, indicative ofsynergisms or antagonisms, are beginning to emerge and require further research. Themagnitude of such deviations cannot be predicted quantitatively. Toxicokineticinteractions are one established cause of deviations from additivity. A notable example ofsuch deviations is the synergism that was observed with a mixture of vinclozolin,prochloraz, finasteride and DEHP with respect to hypospadias and genital malformationsamong male offspring of female rats (Ulla Hass, workshop presentation of unpublisheddata).

4. Cumulative risk assessment – is it necessary?Many experimental studies of mixture effects have been motivated by understandingdeterminants of additivity and predictability. Inevitably, this has meant that chemicalshad to be combined at doses considerably higher than those encountered by the generalpopulation. Two issues need to be addressed to judge the relevance of combinationeffects for risk assessment: Do combination effects occur when chemicals are combinedat low doses? Are the uncertainty factors used to translate apparently safe dose levelsderived from animal experiments into acceptable exposures for humans insufficientlyprotective to take account of mixture effects?4.1 Mixture effects at low doses of mixture componentsCertain experimental mixture studies have been designed to assess whether combinationeffects occur when chemicals are combined at low doses, here defined as beingsufficiently low to be without observable effects when tested on their own (i.e. below thesensitivity of the chosen experiment to be measurable). Often, these doses were in therange of those commonly used to derive estimates of safe human exposures (so-calledpoints of departure, usually no-observed-adverse-effect-levels, NOAELs, or benchmarkdoses). The review by Kortenkamp et al. (2007) summarizes the evidence for endocrinedisrupters and other types of chemicals, and an update was provided by Michael Faust(workshop presentation).For combinations composed of chemicals that interact with the same molecular receptoror molecular target in an organism, there is good evidence that mixture effects can ariseat doses around, or below, points of departure. Considering the main assumptionsunderlying the concept of dose addition, this is to be expected (see 3.1.1).In contrast, theory predicts that mixtures which follow IA should not yield a combinationeffect as long as all components are present at doses associated with zero responses. Thisis widely held to mean that mixtures of dissimilarly acting chemicals are safe, as long asexposure to each component does not exceed its individual point of departure (COT2002, VKM 2008). With reference to the apparent diversity of chemical exposures in the“real world”, IA is taken as the default assessment concept in human toxicology, whenstrict similarity criteria of dose addition appear to be violated or if specific evidence for

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkthe compounds of a given mixture is lacking. Implicitly taking “dissimilar action” or“independent joint action” as the negation of “simple similar action” it is then assumedthat IA must hold, even without further proof that the underlying mechanisms indeedsatisfy any explicit dissimilarity criterion. This is then taken to mean that combinedexposures are without risk as long as all components stay below their points of departure.Consequently, possible mixture effects are considered an irrelevance for chemicals riskassessment.In apparent contradiction to this view, there is good evidence that combinationscomposed of chemicals with diverse modes of action also exhibit mixture effects wheneach component is present at doses equal to, or below points of departure (Kortenkamp etal. 2007, and updates in Michael Faust’s workshop presentation).The flaw in the above line of thinking is two-fold:First, when chemicals cannot be shown to interact with the same molecular targets, itdoes not follow, that they must act in a dissimilar fashion. It is conceivable that diversemodes of action lead to similar adverse outcomes – dissimilar action is not the simplenegation of similar action.Second, points of departure, and particularly NOAELs, are confused with with true zeroeffect levels. Under IA, combination effects cannot arise when the individual responsesof each component in the mixture are zero. With large numbers of chemicals however,even very small individual effects will lead to considerable combined responses. Forexample, 100 chemicals that each produce 0.1% of a maximal effect, are expected toyield a response of 9%, according to IA. However, the resolving power of most testingmethods in regulatory use is far too low to demonstrate such small effects. Far fromsignifying zero effect levels, NOAELs describe a grey zone, where the presence ofeffects can neither be proven, nor ruled out with confidence. NOAELs are frequentlyassociated with effects of between 5 and 10% (Kortenkamp et al. 2007, Scholze andKortenkamp 2007).Taken together, there is good evidence to show that the implicit null-model of manyregulatory assessments, namely, that only the most potent compound determines thetoxicity of the mixture, is usually wrong. Instead, more than one chemical in the mixturecontributes to the observed effects (either according to DA or IA) in contradiction to theregulatory default model of “only the most toxic compound counts”.The demonstration that mixtures of dissimilarly acting chemicals are not without effectwhen they are combined at doses around points of departure, does say little aboutwhether or not risks are present in “real world” exposure settings. The decisive factor forsuch risks to occur lies in the number of chemicals, and their levels: Only if sufficientnumbers of chemicals of sufficient potency and at sufficiently high exposure levels arepresent, are combination effects to be expected. The issue can only be decided on thebasis of better information about relevant combined exposures of human populations andwild life. This information is currently missing, and this knowledge gap presents a majorchallenge to risk assessment.

Expert workshop on combination effects of chemicals, Hornbæk, Denmark4.2 Uncertainty factors in risk assessment and standard setting – do they allow forthe possibility of mixture effects?Although observations of combination effects of endocrine disrupters at low doses havelent urgency to calls to account for such effects in chemicals risk assessment andregulation, the need for doing so is often disputed with the argument that theconventional chemical-by-chemical risk assessment is sufficiently protective. TheUncertainty Factors (UF) usually applied to translate apparently safe dose levels derivedfrom animal experiments into acceptable exposures for humans, so goes the argument,already cover the possibility of combination effects. The issue was examined by MartinScholze (workshop presentation).Uncertainty factors are used in two different ways: Either to assess the health risksassociated with certain chemical exposures by deriving Margins of Exposure (MOE) orMargins of Safety (MOS), or with the aim of establishing recommended health-basedguidance values, such as Acceptable (or Tolerable) Daily Intakes (ADI, TDI), ReferenceDoses (RfD) and such like. Depending on context and goals, they are also referred to asAssessment Factors.The widely used UF of 100 is obtained by multiplication of two factors, one to allow forintra-species sensitivity differences (10), the other for species-species extrapolations fromanimal to human (10). Additional factors may be used to compensate for uncertaintiesdue to lack of information. For example, in the absence of data for chronic toxicity, an(additional) default factor of 10 can be employed. Similarly, if test data do not allow theestimation of a NOAEL, an additional factor of 10 may be brought into play. The variousassessment factors are multiplied, and this can yield a very large overall UF. The largestreported overall UF in USEPA’s Integrated Risk Information System is 10,000. Aspecific factor intended to allow for possible mixture effects is not in use.Nevertheless, the common practice of combining different types of assessment factors bymultiplication has led to the idea that many overall UF’s are overly conservative. Byimplication, this is taken to mean that mixture effects are covered. This idea appears to bebased on a mistaken interpretation of the multiplication rule of probabilities for rareevents. While is it clear that the occurrence of two rare independent events together tendstowards zero, assessment factors cannot be equated with probabilities. A directtranslation of UF’s into probabilities is not possible.There is evidence that the common practice of using a factor of 10 to deal with animal-to-human extrapolations may lead to underestimations of risk. The same applies to the factorof 10 to allow for between-human differences in sensitivity. These considerations forcethe conclusion that an UF of 100 offers insufficient room to allow for mixture effects forall possible realistic mixtures.Finally, the issue of UF’s and mixture effects can be approached from a differentdirection by asking the question: how large would an additional assessment factor have tobe to take account of mixture effects? For a combination of chemicals that follows doseaddition, it can be shown that the RfD’s for each individual chemical would have to bedivided by the number of chemicals that contribute to an overall mixture effect. Forexample, if a combined effect from simultaneous exposure is due to 5 chemicals, then theRfD of every chemical has to be divided by 5, which is equivalent to saying that an

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkadditional assessment factor of 5 is needed to cover mixture effects (NRC 2008).Correspondingly larger factors are needed if more chemicals can be shown to contributeto a common adverse outcome. However, choices about sufficiently protective factorscannot be made without better information about the number of relevant chemicals, theirlevels and potency, and how they contribute to human exposures.To summarize, a specific “mixtures assessment factor” is currently not employed in thetraditional chemical-by-chemical risk assessment, and there is little to suggest thatcommonly used UF are overly protective. There is not much “room” to allow for mixtureeffects.

5. Approaches to Cumulative Risk AssessmentThe practice of Cumulative Risk Assessment (CRA) is furthest developed in the USA,where the US EPA is by far the most important authority for mixtures risk assessmentand regulation. Until recently, a common application of mixtures risk assessment in theUSA was to Superfund waste sites. The Comprehensive Environmental ResponseCompensation and Liability Act (CERCLA) which came into force in 1980 specificallycalls for mixture risk assessment during the evaluation of risks that stem from hazardouswaste sites and chemical accidents. An additional stimulus for CRA was the passage ofthe Food Quality Protection Act (FQPA) in 1996 which required the estimation of healthrisks from combinations of pesticides with a common mode of action, from any exposuresource.Several workshop presentations have dealt with existing approaches and practices ofCRA (presentations by Linda Teuschler, Rolf Altenburger and Henrik Tyle), and oneworkshop aim was to evaluate whether these approaches can be used productively to dealwith endocrine disrupter mixtures.5.1 The grouping of chemicals for the purpose of cumulative risk assessmentCRA begins with the identification of chemicals that should be grouped together andsubjected to joint risk assessment. In Superfund site assessments this is driven byconsiderations of joint exposures. In contrast, CRA for pesticides begins with theidentification of a group of chemicals that are considered to induce a common toxic effectby a common mechanism, a so-called common mechanism group (CMG). The criterionproposed by US EPA (2000) for grouping chemicals for cumulative risk assessment is“toxicological similarity”.Extensive guidance exists about how this should be implemented (US EPA 2000).Pesticides and other chemicals are considered to qualify for inclusion in a CMG whentheir mechanism of toxicity shows similarities in both nature and sequence of majorbiochemical events (workshop presentations by Linda Teuschler and Rolf Altenburger).The use of toxicological similarity based on mechanisms, however, may lead to overlynarrow groupings. For example, organophosphate pesticides and carbamates inhibitacetylcholinesterase, and this is shown to be a relevant step in the manifestation oftoxicity. Because the mechanism of inhibition by carbamates is via carbamylation, and14

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkthat of organophosphates by phosphorylation, and because this is judged to representdifferent molecular mechanisms, the two types of pesticides are not assessed together, butincluded in separate CMGs for the purpose of mixtures risk assessment. Such narrowgroupings ignore that joint effects can also occur from combined exposures with otherthan common mechanisms (workshop presentation by Rolf Altenburger).5.1.1 Grouping for antiandrogensAn exaggerated focus on mechanisms of toxicity may lack plausibility and credibilitywhen it is applied as a grouping criterion for endocrine disrupters. With a recent report bythe National Research Council (NRC) of the US National Academy of Sciences the issuecame to a head with antiandrogens, including phthalates. The NRC advised that acumulative risk assessment should not only consider certain phthalates, but also otherchemicals that could potentially cause the same health effects as phthalates (NRC 2008).It was recommended that phthalates and other chemicals that affect male reproductivedevelopment in animals, including antiandrogens, be considered in the cumulative riskassessment. Solely mechanism-based criteria may lead into a dilemma: Because there aresubtle differences in the precise molecular details by which phthalates can act asendocrine disrupters, not even all antiandrogenic phthalates would be subjected to CRA,when mechanistic considerations are the sole grouping criterion.The NRC therefore recommended a broader based move towards establishing groupingcriteria for phthalates and other antiandrogens. With this type of endocrine disrupter, acase can be made for adopting a physiological approach to analyzing toxic mechanismsof action with respect to similarity or dissimilarity. If it is recognized that the driver ofmale sexual differentiation during development is the effect of androgen action, it isirrelevant whether the hormones’ effects are disrupted by interference with steroidsynthesis, by antagonism of the androgen receptor, or by some other mechanism (forexample, affecting consequences of androgen receptor activation). The resultingbiological effects with all their consequences for male sexual differentiation are similar,although the molecular details of toxic mechanisms - including metabolism, distributionand elimination - differ profoundly in many respects. Judged from such a perspective, afocus on phthalates to the exclusion of other antiandrogens not only would be artificialand lack credibility, but could imply serious underestimation of cumulative risks posedby agents for which there is simultaneous exposure (workshop presentations by Ulla Hassand Andreas Kortenkamp).5.1.2 Thyroid disrupting chemicalsSimilar considerations may apply to the group of thyroid disrupting chemicals whichaffect multiple targets through a variety of mechanisms. In an echo of the situation withantiandrogenic chemical mixtures, the question is: which level of biological complexityshould be used to cumulate joint effects? If an endpoint representative of a specific modeof action is chosen (e.g. variations in T4 levels), then certain chemicals might be left outof a common grouping. On the other hand, if the endpoint chosen for integration is at avery high level of complexity (e.g. changes in cognitive function), not only a very largenumber of chemicals but also a variety of non-chemical stressors will have to be takeninto account. This may become difficult to handle in risk assessment settings (workshoppresentation by Kevin Crofton).

Expert workshop on combination effects of chemicals, Hornbæk, Denmark5.1.3 Dioxin-like endocrine disruptersDioxins and dioxin-like compounds represent a group of endocrine disrupters where keyevents of toxicity are thought to be mediated by binding to the arylhydrocarbon receptor(AhR). These chemicals were first grouped according to descriptors of chemical structure(to include only polychlorinated dibenzo-p-dioxins and –furans, PCDD, PCDF), butinsights into their biological activity led to the incorporation of co-planar PCBs and otherpoly-halogenated polycyclic hydrocarbons (workshop presentation by Martin van denBerg). By using the criterion of AhR activation, polybrominated diphenyl ethers (PBDE)were not included in the group of dioxin-like chemicals. It turned out that pure PBDEwere devoid of AhR activity, and that earlier reports of AhR activation could be ascribedto contamination with dioxin-like chemicals, most importantly polybrominateddibenzodioxins and –furans. The most potent PCDD, 2,3,7,8 TCDD, is selected as thereference chemical, and the potency of all other dioxin-like chemicals is expressed interms of TCDD effect concentrations, so-called TCDD equivalents, with TCDDequivalency factors (TEF) (van den Berg et al 2006). The use of TEF for the assessmentof mixtures of dioxin-like chemicals is an application of the concept of dose addition, andis a widely accepted risk assessment method.5.1.4 Estrogenic chemicalsIn many ways, estrogenic chemicals resemble dioxin-like chemicals: Their activity isthought to be mediated by binding to estrogen receptors (ER alpha or beta), whichsuggests itself as a straightforward grouping criterion (workshop presentation by AndreasKortenkamp). Furthermore, there are reference agents of high potency (17-beta-estradiol,DES) and there is good evidence that mixtures of estrogenic chemicals follow doseaddition when the assessment is based on events relatively close to receptor activation(Kortenkamp 2007). Consequently, it has been suggested that this group of endocrinedisrupters should be assessed just like dioxin-like chemicals, by using the toxicityequivalency concept. However, this suggestion has been called into question by Safe,with reference to the complexity of estrogen signaling (discussed in Kortenkamp 2007).Nearly 20 years ago, evidence has emerged that ER activation is possible without bindingto the binding pocket of the steroid hormone, e.g. by phosphorylation through activationby growth factors. This opens the possibility that estrogen action can be substantiallymodulated by chemicals interfering with other phosphorylation events. Should suchagents be subjected to CRA with estrogenic chemicals? Furthermore, more research isneeded to elucidate the toxicological relevance of ER activation. Although chemicalssuch as DES are potent disrupters of male and female sexual differentiation, it remains tobe seen whether these effects are mediated by ER activation. Similarly, the mechanismsby which estrogens play a role in breast cancer are not entirely resolved.Considerations of mode of action as a grouping criterion are often of little use inecotoxicological mixtures risk assessment, because each chemical usually exhibitsmultiple modes of action. The solution to this problem is to take account of sensitivitydifferences in various receptors and species (Leo Posthuma, workshop presentation).

Expert workshop on combination effects of chemicals, Hornbæk, Denmark5.2 Mixtures risk assessment methodsThe application of mixtures risk assessment methods requires clarity about the goal of theassessment. The aim can be to arrive at a risk estimate, an estimation of safe levels, ofmargins of exposure, or can consist in ways to prioritize certain mixtures (LindaTeuschler and Leo Posthuma, workshop presentations). Estimations of safe levels ormargins of exposure may be based on worst-case-assumptions, but the prioritization ofmixtures (or affected sites) has to rely on fairly accurate quantitations of risk.Considering that experimental studies with endocrine disrupters showed that doseaddition is a useful concept for the approximation of combination effects, component-based methods derived from dose addition suggest themselves as risk assessmentapproaches. These include the Hazard Index (HI), Point of Departure Index (PODI) andthe TEQ concept (Linda Teuschler and Henrik Tyle, workshop presentations).All these methods require dose-response information of mixture components as inputvalues. The HI sums up ratios of exposure levels and reference doses over chemicals. Thereference doses can be arrived at by utilizing different UF for each mixture component. Ifthis is perceived to be a problem, the PODI method can be used. PODI is based not onreference doses, but on points of departure (NOAELs, benchmark doses). Extrapolationissues (e.g. animal to human) are then dealt with by using one overall UF. Finally, theTEQ concept is predicated on the choice of a reference chemical and requires paralleldose-response curves for all components. Both these requirements are often not met byendocrine disrupters, but the method has been validated for dioxin-like endocrinedisrupters.5.2.1 TieringDepending on the quality of the data that are available for CRA (data poor or data rich),tiering methods might be very productive to explore the problem, and refine (with moresophisticated models and associated supporting data) when needed (Leo Posthuma,workshop presentation). At the lowest tier (tier 0), it may become apparent that thesituation to be evaluated does in fact not present an issue for mixtures risk assessment. Inthe next higher tier (tier 1), termed “simple generic”, data about mixed exposures may notbe present, but it may be deemed desirable to safeguard against the possibility of jointeffects by adopting a specific mixtures assessment factor. In tier 2, “moderately simplegeneric”, sufficient data may be available to warrant the assumption of dose additionthroughout, in which case variants of this concept could be applied, even thoughindependent action may produce less conservative estimates. In a quite data rich situation(tier 3, “complex specific”) sufficient information about various modes of action may beavailable, such that mixed mixtures assessment models (DA within groups of compoundsperceived to follow simple similar action, followed by IA across groups) can be applied.Finally, in the highest tier 4 (“highly specific”) it would be possible to address both issuesof modes of action and differences in the vulnerability of various species or riskreceptors.In the light of the data situation typical for many endocrine disrupters, it would appearthat assessments at tier 1 and tier 2 are currently possible.

Expert workshop on combination effects of chemicals, Hornbæk, Denmark5.2.2 ValidationWith the aim of putting CRA methods on a sound footing, it is important to seeksituations where the outcome of specific assessments can be validated. While this isachievable in ecotoxicology (presentation Leo Posthuma), the situation is much morecomplicated in the arena of human toxicology.5.3 Regulation and risk managementCRA for endocrine disrupters and other chemicals can yield important stimuli forregulation and risk management, by providing the basis for a procedure of relativeranking, e.g. according to the most potent chemicals. This would offer the possibility ofstrictly regulating, or even eliminating those chemicals that are shown to have thegreatest impact on a combination effect. Other rankings could be performed in terms ofthe most problematic exposure settings, or the most vulnerable population subgroups(Leo Posthuma, workshop presentation).

6. Consensus formulation and recommendationsThe workshop participants reached a consensus on a number of specific issues relevant toCRA of endocrine disrupters. The participants also made certain recommendationsconcerning risk assessment methods, research needs, and legislative requirements.6.1 Mixtures risk assessment is necessaryIn view of the evidence about mixture effects at low experimental doses (see 4.1) and theuncertainty of commonly employed UF in single-chemical risk assessment (see 4.2) adisregard for combination effects was considered undesirable and not in line withcurrently available empirical evidence. Any CRA method, even one that employs thenarrowest possible toxicological grouping criteria, was deemed to represent animprovement compared to the current pre-occupation of conventional risk assessmentwith chemical-by-chemical approaches. Moreover, an extended look at simultaneous orsequential exposure issues was deemed crucial, to add to the classical toxicologicalapproaches.6.2 The assumption of independent action as a default for “real world” mixtures isnot tenableAs discussed in 4.1, the absence of proof of “similarity” in the mode of action of mixturecomponents cannot be taken to indicate applicability of IA as a default, with the implicitassumption that combination effects are not to be expected if all chemicals are present atdoses below their individual points of departure or NOAELs. The workshop participantsrecognized that the available empirical data do not support this widely held view. Instead,there is good evidence that mixtures that follow IA exert effects even when all mixturecomponents are present at doses below their NOAELs.

Expert workshop on combination effects of chemicals, Hornbæk, Denmark6.3 The application of dose addition is recommended as a default, until evidence tothe contrary appearsThe empirical evidence with endocrine disrupter mixtures (see 3.3) shows that DA yieldsreasonable approximations of observed combination effects. There are many exampleswhere IA has produced underestimations of observed joint effects. Crucially, no casecould be identified, where IA afforded a more conservative mixture effect prediction thatwas at the same time in agreement with the experimental mixture effects. It isconceivable that such evidence appears in the future, but until this is the case, DA wasrecommended as the default assessment method, irrespective of presumed modes ofaction. This modus operandi has the additional advantage of requiring fewer data than thealternative concept of IA, with the consequence that it can serve as lower-tier approach inmany circumstances.6.4 Steps towards CRA: criteria for the grouping of chemicals, assessment methodsGrouping criteria that are driven exclusively by thoughts about mechanisms or the keyevents for a mode of action were seen as problematic by the workshop participants (see5.1), and it was recommended that grouping should be dissociated from mechanisticconsiderations. For risk assessment, phenomenological grouping criteria, based oncommon adverse health outcomes, were seen as a more useful starting point forgroupings. Nevertheless, it was recognised that toxic effects become less specific for theinitiating event, as one moves further down-stream of an effector chain. This loss ofspecificity may lead to the inclusion of an ever wider array of chemicals into a groupingfor CRA, ultimately blurring all distinctions, with the need to include all chemicals.However, this was not perceived to be a critical problem for endocrine disruptingchemicals.Another useful criterion for groupings is the likelihood with which simultaneousexposures to several chemicals occurs.A tiered assessment, depending on the extent and quality of existing data about hazardsand exposures is recommended. For example, to alleviate concerns about mixture effects,it would be possible to adopt a specific assessment factor in the traditional chemical-by-chemical risk assessment, even without any further data. In more data-rich situations, it isfeasible to utilize applications of the dose addition concept to define margins of exposureor other indicators of risk.6.5 CRA for endocrine disrupters, although feasible, is hampered by important datagapsDue to significant experimental advances in the last five years, determinants of additivemixture effects of classes of endocrine disrupters are now quite well understood. Theprospect of CRA for endocrine disrupters is limited by incomplete information aboutrelevant exposure scenarios. This is particularly critical for human risk assessment: it isnot even possible to say with confidence whether there are only a few chemicals thatcontribute significantly to an overall mixture effect, or whether the number of relevantchemicals is likely to be high. Better knowledge about this aspect of the problem would19

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkhave an enormous impact on the prospects of CRA. The issue can only be resolvedthrough dedicated mixtures exposure assessment approaches, where scores of chemicalsare measured in one and the same sample. This would also provide information about thefeasibility of using certain index chemicals as surrogates for exposure measurements.Furthermore, it was recommended to identify exposure “hot spots” with the aim of usingthose for targeted monitoring (with associated exposure ‘cold spots’ as points ofreference for interpretation).Another challenge concerns the issue of dose metrics. The usefulness of data from animalexperiments would be enhanced greatly if the internal tissue levels resulting fromexperimental exposures were known. This would enable a read-across to readilyaccessible data about human tissue levels.Further research needs are in the following areas:The joint effects of different classes of endocrine disrupters need to be evaluated, and abetter understanding of hormone systems other than estrogens, androgens and thyroidhormones is urgently required.Finally, determinants that lead to synergisms between endocrine disrupters need to beinvestigated.6.6 A better legislative basis for CRA is needed in EuropeWithout the legal mandates laid down in the US American CERCLA and FQPA,cumulative risk assessment would not have been implemented in the USA. With theexception of the recent changes in European pesticides regulations, where mixture riskassessment is mandated, comparative legal frameworks that clearly address CRA docurrently not exist in Europe. In REACH for example, CRA for multiple chemicals frommultiple sources, routes and pathways is only addressed to a very limited extent in thecurrent guidance. Other relevant European legislations do not contain a mandate for CRAfor multiple chemicals from multiple sources, routes and pathways.Development of a comprehensive implementation of CRA should be given seriousconsideration in all relevant legislation and guidance dealing with chemicals safetyassessment and the establishment of safe emission and exposure levels. It is essential toassess the scope of existing laws and guidance in order to define better whether existingregulation can be amended to accommodate CRA, or whether tailor-made regulationsneed to be developed.6.7 PrioritisationThe workshop participants were asked to distill their recommendations into a few mainpoints and to prioritize. Consensus on the following was reached:•CRA for endocrine disrupters can start immediately – important informationnecessary to make decisions about groupings of chemicals to be subjected tomixture risk assessment is available.

7. ReferencesBackhaus T, Scholze M, Grimme LH. 2000. The single substance and mixture toxicity ofquinolones to the bioluminescent bacteriumVibrio fischeri.Aquat Toxicol 49:49-61.Berenbaum MC. 1985, "The expected effect of a combination of agents: the generalsolution",Journal of Theoretical Biology,114: 413-431.Bliss CI. 1939. The toxicity of poisons applied jointly.Ann Appl Biol26:585-615.COT (Committee on Toxicity of Chemicals in Food, Consumer Products and theEnvironment). 2002. Risk assessment of mixtures of pesticides and similarsubstances. Her Majesty’s Stationary Office, London, United Kingdom. Available:http://www.food.gov.uk/science/ouradvisors/toxicity/COTwg/wigramp/[accessed 7September 2005].Drescher K, Boedeker W. 1995. Concepts for the assessment of combined effects ofsubstances: the relationship between concentration addition and independent action.Biometrics51:716-730.Faust M. et al. 2003, "Joint algal toxicity of 16 dissimilarly acting chemicals ispredictable by the concept of independent action.",Aquatic Toxicology,63: 43-63.Kortenkamp A. 2007, "Ten years of mixing cocktails: a review of combination effects ofendocrine-disrupting chemicals",Environ.Health Perspect.,115 (Suppl 1): 98-105.Kortenkamp A, Faust M, Scholze M,Backhaus T. 2007, "Low-level exposure to multiplechemicals: reason for human health concerns?",Environ.Health Perspect.,115(Suppl 1):106-114.Loewe S, Muischnek H. 1926. Über Kombinationswirkungen. 1. Mitteilung: Hilfsmittelder Fragestellung [in German].Naunyn-Schmiedebergs Arch Exp PatholPharmakol114:313-326.NRC, 2008. Phthalates Cumulative Risk Assessment – The Tasks Ahead. Committee onPhthalates Health Risks, National Research Council, National Academy ofSciences, Board on Environmental Science and Technology, National AcademyPress, Washington, DC.Rider CV, Furr J, Wilson VS, Gray LE., Jr. 2008, "A mixture of seven antiandrogensinduces reproductive malformations in rats",Int. J. Androl,31: 249-262.Scholze M, Kortenkamp A. 2007. Statistical power considerations show the endocrinedisrupter low dose issue in a new light.Environ. Health Perspect.115 (Suppl.1):84-90.Suter GW, Cormier SM. 2008. What is meant by risk-based environmental qualitycriteria.Integrated Environ Assess Monitoring4: 486-489.U.S. EPA. 1986. Guidelines for health risk assessment of chemical mixtures. Fed. Reg.51(185):34014-34025.

Expert workshop on combination effects of chemicals, Hornbæk, DenmarkU.S. EPA. 1989. Risk assessment guidance for superfund. Vol. 1. Human healthevaluation manual (Part A). EPA/540/1-89/002. Washington, DC:U.S.Environmental Protection Agency.U.S. EPA. 2000. Supplementary guidance for conducting health risk assessment ofchemical mixtures. EPA/630/R-00/002. Washington, DC:U.S. EnvironmentalProtection Agency.Van den Berg M, Birnbaum L, Bosveld ATC, Brunstrom B, Cook P, Feeley M, et al.1998. Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans andwildlife.Environ Health Perspect106:775-792.Van den Berg, M., Birnbaum, L.S., Denison, M., De Vito, M., Farland, W., Feeley, M.,Fiedler, H., Hakansson, H., Hanberg, A., Haws, L., Rose, M., Safe, S., Schrenk, D.,Tohyama, C., Tritscher, A., Tuomisto, J., Tysklind, M., Walker, N., and Peterson,R.E., 2006. The 2005 World Health Organization Reevaluation of Human andMammalian Toxic Equivalency Factors for Dioxins and Dioxin-Like Compounds.Toxicol. Sci.93: 223–241.VKM. 2008. Combined toxic effects of multiple chemical exposures. Opinion of theScientific Steering Committee of the Norwegian Scientific Committee for FoodSafety. Oslo. ISBN (printed version) 978-82-8082-232-1

Expert workshop on combination effects of chemicals, Hornbæk, Denmarkassessed? What are criteria for grouping these substances for purposes of mixtures riskassessment?15:45Ulla HassCombination effects of phthalates and other anti-androgens aftergestational exposure

Resource:

Hass et al. 2007 EHP 115, Suppl 1 : 122, Metzdorff etal. 2007 Toxicol Sci 98 : 87, Christiansen et al. 2008 Int. J.Androl. 31: 24116:1516:30DiscussionAndreas KortenkampWhich chemicals should be grouped to protect against combinationeffects resulting in disruption of male sexual differentiation? – adiscussion of grouping criteria

Resource:

Summary chapter of US NRC report “Cumulative riskassessment for phthalates – the tasks ahead”17:0017:15DiscussionLinda TeuschlerAn overview of chemical mixtures risk assessment methods

Resource:

Teuschler 2007 TAP 223: 139Having discussed the specifics of antiandrogen mixtures in some detail, this presentationis intended to broaden the discussion and will summarize the methods that have beenused to group other substances for the purpose of mixtures risk assessment. Is it possibleto derive generally applicable criteria?17:4518:00DiscussionDrinks and dinner

Thursday, 29 January 2009Session 3: The basis of combined risk assessment for other classes ofendocrine disrupters and other chemicalsThe following series of talks will consider topics for mixture risk assessment relevant toother endocrine disrupting chemicals, such as: What are effect outcomes or mechanismson which mixtures risk assessment should be based? What is the evidence forcombination effects?9:00Kevin CroftonEffect profiles of thyroid-disrupting chemicals and experimentalevidence of mixture effects

Resource:

Crofton 2008 IJA, Crofton et al. 2005 EHPDiscussion9:30Martin van den BergDioxins, PCBs and related chemicals – an update on the TEFapproach

Resource:

Van den Berg et al. 2006 Tox Sci 93 : 223Discussion10:00Andreas KortenkampEstrogens and estrogen-like chemicals – an update on combinedeffects

Resource:

Kortenkamp 2007EHP 115 Suppl 1: 98Discussion10:3011:00Coffee breakRolf AltenburgerA brief overview of other efforts of mixtures risk assessment:organophosphates, carbamates, chloroacetanilides, triazines…

Resource:

EPA guidance documents

Session 4: From mixtures risk assessment to regulationThe scene is set for a more general treatment of the mixtures risk assessment, relevant toother groups of chemicals. The session begins with a brief summary of approaches tomixtures regulation, considers practice in ecotoxicology, and what can be derived forhuman toxicology and ends with an analysis of uncertainty factors and their suitabilityfor covering mixture effects.14:00Henrik TyleSynopsis of approaches to mixtures regulation (top n, PODI, HI,TEF, relative potency factors, etc)Resource: VKM report 2008, p 38 – 51, Feron et al 2004, ETAP18, 21514:30Leo PosthumaPractical approaches in ecotoxicological mixture risk assessment insupport of urgent policy questionsMartin ScholzeUncertainty factors in standard setting – are mixture effectscovered?Coffee breakGeneral discussion – focus: can existing chemicals regulationbe modified to take account of mixtures effects?Break-out group: Formulation of theses and summary

Resources references(in the order of the talks)

Michael Faust

Kortenkamp, A., Faust, M., Scholze, M., & Backhaus, T. 2007, "Low-level exposure tomultiple chemicals: reason for human health concerns?",Environ.Health Perspect.,vol.115 Suppl 1, pp. 106-114.

Ulla Hass

Hass, U., Scholze, M., Christiansen, S., Dalgaard, M., Vinggaard, A. M., Axelstad, M.,Metzdorff, S. B., & Kortenkamp, A. 2007, "Combined exposure to anti-androgensexacerbates disruption of sexual differentiation in the rat",Environ.Health Perspect.,vol.115 Suppl 1, pp. 122-128.Metzdorff, S. B., Dalgaard, M., Christiansen, S., Axelstad, M., Hass, U., Kiersgaard, M.K., Scholze, M., Kortenkamp, A., & Vinggaard, A. M. 2007, "Dysgenesis andhistological changes of genitals and perturbations of gene expression in male rats after inutero exposure to antiandrogen mixtures",Toxicological Sciences,vol. 98, no. 1, pp. 87-98.Christiansen, S., M. Scholze, M. Axelstad, J. Boberg, A. Kortenkamp, & U. Hass. 2008.Combined exposure to anti-androgens causes markedly increased frequencies ofhypospadias in the rat.Int. J. Androl.Vol 31, no 2, pp.241-248.

Andreas Kortenkamp

NRC, 2008. Phthalates Cumulative Risk Assessment – The Tasks Ahead. Committee onPhthalates Health Risks, National Research Council, National Academy of Sciences,Board on Environmental Science and Technology, National Academy Press, Washington,DC.

Linda Teuschler

Teuschler, L.K. 2007, Deciding which chemical mixtures risk assessment methods workbest for what mixtures.Toxicology and Applied Pharmacology,vol 223, pp 139-147.

Kevin Crofton

Crofton, K. M., Craft, E. S., Hedge, J. M., Gennings, C., Simmons, J. E., Carchman, R.A., Carter, W. H., & Devito, M. J. 2005, "Thyroid-hormone-disrupting chemicals:Evidence for dose-dependent additivity or synergism",Environmental HealthPerspectives,vol. 113, no. 11, pp. 1549-1554.

Martin van den Berg

Van den Berg, M., Birnbaum, L.S., Denison, M., De Vito, M., Farland, W., Feeley, M.,Fiedler, H., Hakansson, H., Hanberg, A., Haws, L., Rose, M., Safe, S., Schrenk, D.,Tohyama, C., Tritscher, A., Tuomisto, J., Tysklind, M., Walker, N., & Peterson, R.E.,2006. The 2005 World Health Organization Reevaluation of Human and MammalianToxic Equivalency Factors for Dioxins and Dioxin-Like Compounds.Toxicol. Sci.vol93 no. 2, pp 223–241

Andreas Kortenkamp

Kortenkamp, A. 2007, "Ten years of mixing cocktails: a review of combination effects ofendocrine-disrupting chemicals",Environ.Health Perspect.,vol. 115 Suppl 1, pp. 98-105.

Rolf Altenburger

US EPA, 1999. Guidance for Indentifying Pesticide Chemicals and Other Substances thatHave a Common Mechanism of Toxicity. Office of Pesticide Programs, USEnvironmental Protection Agency, Washington, DCUS EPA 2006a. Cumulative Risk from Triazine Pesticides. Office of Pesticide Programs,US Environmental Protection Agency, Washington, DC., March 2006US EPA 2006b. Cumulative Risk from Chloroacetanilide Pesticides. Office of PesticidePrograms, US Environmental Protection Agency, Washington, DC., March 2006

Henrik Tyle

VKM. 2008. Combined toxic effects of multiple chemical exposures. Opinion of theScientific Steering Committee of the Norwegian Scientific Committee for Food Safety.Oslo. ISBN (printed version) 978-82-8082-232-1Feron, V.J., van Vliet, P.W. & Notten, R.F., 2004. Exposure to combinations ofsubstances: a system for assessing health risks.Environ Toxicol Pharmacolvol 18, pp.215-222.