Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Current Practice in Risk Assessment and Cumulative Risk Assessment Chapter 2 summarizes the evidence on human exposure to phthalates and demonstrates that there is ample evidence of simultaneous exposure of most of or all the U.S. population to multiple phthalates. Chapter 3 examines the toxic- ity, as seen primarily in laboratory animal models, of individual phthalates and of other agents that produce effects similar to those seen on exposure to individ- ual phthalates. The exposure and toxicity information clearly indicates that some sort of cumulative risk assessment is required in examining phthalate exposure. To place the discussions of Chapter 5 in context, it is useful to observe what is currently done when risks posed by multiple chemical exposures are evaluated with standard techniques according to guidance of the U.S. Environmental Pro- tection Agency (EPA) and to examine how the guidance has evolved. CURRENT RISK-ASSESSMENT APPROACHES AND PRACTICES The most extensive and detailed guidance on typical risk assessments is in the Risk Assessment Guidance for Superfund (RAGS), particularly Volume I, Human Health Evaluation Manual (Part A) (EPA 1989a) and later published guidance supporting RAGS. This chapter features a description of how risk as- sessment is performed with RAGS because it (as intended according to its statement of purpose) tends to inform risk-assessment practice in other EPA programs and under other regulatory authorities. Additional guidance documents are cited where needed. EPA guidance on cumulative exposure and risk has evolved over the last couple of decades, and the relevant developments are dis- cussed near the conclusion of this chapter in the section âThe Evolution of Guidance on Cumulative Risk Assessment.â The chapter concludes with a summary of recent cumulative exposure and risk evaluations. The reason for considering RAGS and the actual procedures that are used in the field is to emphasize that what is done in site-specific risk assessments (for example, at Superfund sites) is distinct from the approaches and procedures 68
Current Practice in Risk Assessment and Cumulative Risk Assessment 69 used in setting standards or guidelines for individual chemicals. Although both draw heavily on toxicity assessmentsâfor example, as appear on or are used by EPAâs Integrated Risk Information System (IRIS) web siteâthe application of the toxicity assessments typically differs considerably between the two. Site- specific risk assessments are often concerned with simultaneous evaluation of multiple chemicals, multiple pathways of exposure, multiple routes of exposure, and multiple receptors. Standard-setting or guideline-setting generally evaluates at one time single chemicals, single routes of exposure, and single receptors, although there are exceptions, such as disinfection byproducts in drinking water. The committeeâs task of evaluating the potential for a cumulative risk as- sessment of phthalates has to take into account that such cumulative assessments are commonly performed already, and any recommendations of the committee should be compared with the current EPA approach as described in RAGS and related guidance. Accordingly, the following sections discuss what is typically required in the exposure-assessment, toxicity-assessment, and risk- characterization parts of a risk assessment. The approaches are evaluated for what they imply about cumulative assessment of phthalates in various EPA pro- grams, such as those involving Superfund, air toxics, and drinking water. Exposure Assessment The exposure-assessment component of a risk assessment of hazardous chemicals according to RAGS (EPA 1989a) requires evaluation of exposure of all the relevant, although not personally identified, people (âreceptorsâ) to all the relevant chemicals through all the relevant pathways by all the relevant routes of exposure for all relevant periods. The products of exposure assessment are esti- mates of exposure of defined receptors to each chemical disaggregated by peri- ods and exposure pathways. This section provides an idealized general descrip- tion, not a critical review, of the current practice of exposure assessment. Persons Whose Exposure Is Quantified The relevant receptors to evaluate are typically intended to be persons who experience the âreasonable maximum exposureâ (RME) and persons who ex- perience âcentral-tendencyâ (CT) exposures. The RME is the highest exposure that is expected to occur (EPA 1989a, 1992), and EPA (2001) advises that risk managers using probabilistic risk assessment should select the RME from the upper end of the range of risk estimates, âgenerally between the 90th and 99.9th percentilesâ (EPA 2001). Later discussion focuses on persons who experience the RME because their exposure usually forms the basis of EPA decision- making (CT estimates may be needed for some pathways of the RME, as de- scribed below).
70 Phthalates and Cumulative Risk Assessment: The Tasks Ahead Chemicals Warranting Quantitative Dose Estimation The relevant chemicals to evaluate in an exposure assessment are those which pass an initial screening evaluation that is used to eliminate chemicals that are clearly of no concern. The evaluation will typically first examine any available observations for frequency of occurrence and concentrations of chemi- cals in whatever physical media have been examined; this eliminates chemicals that occur very rarely and at concentrations much lower than risk-based screen- ing valuesâprecalculated values that, if they were carried through a risk as- sessment, would result in risk estimates small enough to be ignored. Where the only concern is increments of exposure above background, chemicals whose concentrations are similar to background may also be eliminated from further consideration. Further screening may be performed to demonstrate that even worst-case exposures (based on upper-bound estimates of exposure) present no hazard. An exposure assessment is typically applied for many chemicals, although usually the nature of the expected major contamination is known to some de- gree. For example, the initial list of chemicals to be evaluated in a typical site risk assessment is usually the Contract Laboratory Program Target Compound and Target Analyte List (TCP/TAL, see EPA 2008a), combined with any site- specific chemicals known to be present and to have potential toxicity. The TCP/TAL (as of May 2008) includes 52 volatile chemicals, 30 pesticides and Aroclors, 23 metals, cyanide, and 67 semivolatile chemicals. The semivolatile chemicals include six phthalates: DMP, DEP, DBP, BBP, DEHP, and DOP. For an exposure assessment performed for a risk assessment at a contami- nated siteâfor example, a Superfund site or a site evaluated under similar state programsâenvironmental samples will often be tested for all chemicals on the TCP/TAL or similar lists, augmented where necessary. An initial screening for the full list of chemicals may be performed on a small number of samples cho- sen from areas thought to be most contaminated (for example, because of visual observation of soil staining, according to known locations of potentially con- taminating processes, or on the basis of on-site screening with vapor detectors or conductivity measurements), and chemicals that are not detected may be dropped from the analytic sample suite. Later samples may be analyzed for a smaller list of chemicals. As discussed above, not all the chemicals analyzed will be evaluated through all parts of the exposure and risk assessment; applica- tion of screening approaches may allow chemicals to be dropped from some exposure pathways or for some receptors. For some sites or situations, there will be evaluation of special compounds not included in the lists described or special analyses of the compounds listed. For example, where contamination by poly- chlorinated dibenzo-p-dioxins (PCDDs) or polychlorinated dibenzo-p-furans (PCDFs) is suspected or found in an initial screening, analyses of various PCDD or PCDF congeners may be conducted.
Current Practice in Risk Assessment and Cumulative Risk Assessment 71 Exposure Pathways and Periods Evaluated Exposure assessment should take account of all the exposure pathways that can occur for any person. The relevant pathways included are all those by which some chemical may travel and cause exposure to the chosen receptors (that is, complete pathways). The relevant routes of exposure (ingestion, inhala- tion, and dermal contact) are all that may occur at the end of any particular pathway; in special circumstances, other routes, such as injection or transmuco- sal absorption, might have to be considered. The relevant periods depend on the toxic characteristics of the chemicals evaluated and on the timing and pattern of exposure but typically are handled by estimating exposure averaged over fixed periods for various locations and characteristics of receptorsâsuch as age, sus- ceptibility, and habits. Typically, assessments will evaluate acute exposure (from instantaneous to a few days long), subchronic exposure (from a few days to about 7 years), and chronic exposure (extending to a lifetime). Total Doses Estimated for Each Receptor For each pathway, the exposures of the receptor who experiences the RME are obtained by using procedures that result in estimates at the upper end of likely exposures, but the extent of any underestimation or overestimation is not generally known. More complex procedures, such as probabilistic methods, may be used to obtain better estimates of explicit percentiles of the exposure distribu- tion. If it is determined that combined exposure (from multiple pathways) can occur, receptors who experience the RME are defined for combinations, and upper-end estimates of combined exposures are obtained by summing suitable combinations of estimates for each pathway. Such combinations may involve upper-end estimates for one or more pathways and average estimates for others. The aim is to obtain exposure estimates that are at the upper end of the actual or potential exposure. The result is a total dose estimate for each receptor, disag- gregated by chemical, route of exposure, and period. Toxicity Assessment As in the preceding section, this section provides an idealized general de- scription, not a critical review, of the current practice of toxicity assessment. General Approach The practical and most commonly adopted approach to toxicity assess- ment in EPA risk assessments is to obtain toxicity values from the EPA IRIS database for chronic oral reference doses (RfDs), chronic inhalation reference concentrations (RfCs), cancer classification, ingestion and inhalation cancer
72 Phthalates and Cumulative Risk Assessment: The Tasks Ahead slope factors (CSFs for lifetime exposure), and inhalation and ingestion unit risks (URs for lifetime exposure). (See Box 4-1 for how those quantities are defined by EPA on its IRIS web site.1) In some cases, such as that of vinyl chlo- ride, IRIS provides modifications of the values, for example, separate estimates of oral CSF or UR for continuous lifetime exposure during adulthood and for continuous lifetime exposure from birth. BOX 4-1 EPA Definitions for Toxicity Values Cancer Evaluations Cancer slope factor (CSF): An upper bound, approximating a 95% confidence limit, on the increased cancer risk from a lifetime exposure to an agent. This estimate, usually expressed in units of proportion (of a population) affected per mg/kg-day, is generally reserved for use in the low- dose region of the dose-response relationship, that is, for exposures corresponding to risks less than 1 in 100. Unit risk (UR): The upper-bound excess lifetime cancer risk estimated to result from continuous exposure to an agent at a concentration of 1 Âµg/L in water, or 1 Âµg/m3 in air. The interpretation of unit risk would be as follows: if unit risk = 2 x 10-6 per Âµg/L, 2 excess cancer cases (upper bound estimate) are expected to develop per 1,000,000 people if exposed daily for a lifetime to 1 Âµg of the chemical in 1 liter of drinking water. Noncancer Evaluations Reference concentration (RfC): An estimate (with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. It can be derived from a NOAEL [no-observed-adverse-effect level], LOAEL [lowest observed-adverse-effect level], or benchmark concentration, with uncertainty factors generally applied to reflect limitations of the data used. Reference dose (RfD): An estimate (with uncertainty spanning perhaps an order of magnitude) of a daily oral exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. It can be derived from a NOAEL, LOAEL, or benchmark dose, with uncertainty factors generally applied to reflect limitations of the data used. Source: EPA 2008b. 1 The committee has not examined whether the values used by EPA meet the defini- tions.
Current Practice in Risk Assessment and Cumulative Risk Assessment 73 IRIS is at the top of EPAâs recommended three-tier hierarchy of sources for toxicity values for use at Superfund sites (EPA 1993a, 2003a); more broadly, IRIS values support EPA policy-making activities (EPA 2008c). When IRIS does not provide toxicity values or when toxicity values are needed for circum- stances not typically provided for in IRIS (for example, for evaluation of sub- chronic or acute exposures2), the recommended hierarchy of sources is searched sequentially for suitable values. However, EPA recognizes that the hierarchy does not address situations where new toxicity information is brought to its at- tention. Therefore, although in practice risk assessments typically incorporate previously developed toxicity values, especially IRIS values, new information could result in the development and application of toxicity values other than those in EPAâs hierarchy. The derivation of toxicity values for non-EPA risk assessments, such as those performed for or by state agencies, typically follows the same patterns as for EPAâs risk assessments for Superfund sites. Toxicity values are typically predefined by a state agency with jurisdiction, usually by reference to a hierar- chy of sources of toxicity values prepared by other suitably authoritative sources, although the hierarchy may differ from EPAâs and from state to state. The Environmental Protection Agencyâs IRIS Process The output of the IRIS process is a set of toxicity values that can be used in site-specific risk assessment, such as for Superfund sites; product-specific risk assessments, such as those for consumer products; media-specific risk assessments, such as for drinking-water standards; and other applications of risk assessment. Those conducting the risk assessments must confirm the relevance of IRIS values for the chemical species, exposure pathway, exposure timeframe (nearly all toxicity values on IRIS apply to the evaluation of chronic exposures), and population under evaluation (for example, in case the population might have increased susceptibility with respect to life stage, disease status, or genetic predisposition that is not already accounted for in development of the toxicity value). The IRIS database on a chemical contains the toxicity values and brief summaries of toxicity data and other information that support them. Since 1997, the database summaries have been supplemented by detailed toxicologic reviews that undergo an independent expert peer review, including the opportunity for public review and comment. Toxicologic reviews summarize a chemicalâs properties, toxicokinetics, pharmacokinetic modeling where available, hazard identification based on epidemiologic studies, animal studies, in vivo and in vitro assays, and mechanism-of-action and dose-response data and culminate in quantitative recommendations for toxicity values when sufficient data are 2 A recent exceptional case that provides values for subchronic and acute exposures is that of 1,1,1-trichloroethane (see EPA 2007a).
74 Phthalates and Cumulative Risk Assessment: The Tasks Ahead available to support them. In conducting toxicologic reviews, EPA uses relevant guidance that includes evaluation of the array of possible health outcomes, such as cancer, neurotoxicity, developmental toxicity, and reproductive toxicity. Toxicity Values Currently Available for Phthalates Table 4-1 summarizes toxicity values currently available for phthalates in the EPA hierarchy of sources.3 IRIS provides a limited set of toxicity values for five phthalates. As discussed earlier, IRIS values make up the highest tier in EPAâs hierarchy of toxicity values (EPA 2003a), and EPA generally favors their use, when available, over lower-tier toxicity values. Provisional peer-reviewed toxicity values (PPRTVs) make up the second tier of toxicity values and are developed by the Superfund Health Risk Technical Support Center (STSC). The STSC has assigned a PPRTV to BBP and a âscreening valueâ to DMP, which are available with supporting documentation internally to EPA and on request to registered users. The third tier of EPAâs hierarchy of toxicity values, which is a catch-all for âother toxicity values,â includes California Environmental Protection Agency Maximum Allowable Dose Levels (MADLs) and Agency for Toxic Substances and Disease Registry (ATSDR) minimal risk levels (MRLs). California has established MADLs for two phthalates, DEHP and DBP; however, the value for DBP is not in the database identified in the EPA procedure (EPA 2003a) for locating other toxicity values, so it has not been included in Table 4-1. ATSDR has established MRLs based on noncancer end points for four phthalates: DEHP (reproduction end point), DBP (developmental end point), DOP (hepatic end point), and DEP (reproductive and hepatic end points). Only the MRL for DEHP applies to the evaluation of chronic exposures, defined by ATSDR as lasting over 365 days. The other MRLs apply to acute exposure (1-7 days) and intermediate exposure (7-364 days). Only three of the seven phthalates known to cause phthalate syndrome in rats (see Table 3-3) have toxicity values in this hierarchy. Furthermore, the val- ues are based on nonreproductive toxicities with the exception of the ATSDR MRL and the California MADLs for DEHP, and two others (DMP and DEP) are listed that have not been associated with phthalate syndrome. The screening value for DMP developed by EPAâs STSC is based on a lowest observed- adverse-effect level associated with increased absolute and relative liver weight and decreased serum and testicular testosterone in weanling male rats. Despite noting the observed lack of adverse effects of DMP on reproductive outcomes or fetal development, the authors of the screening value concluded that âexposure 3 The committee notes that it was not charged with reviewing the basis or adequacy of the values reported in Table 4-1; the committee is simply reporting the current toxicity values for phthalates.
TABLE 4-1 Summary of EPAâs Toxicity Values for Phthalatesa Phthalateb Chronic Oral Reference Chronic Inhalation Reference CAS No. Dose (mg/kg-d) Concentration (mg/m3) Oral Slope Factor (mg/kg-d)-1 Inhalation Unit Risk (Âµg/m3)-1 DMP CAS Not available (3/1/1994)c Not available (10/1/1990); EPA Not available; âDâânot Not available; âDâânot classifiable 131-11-3 contractor updated review classifiable (2/1/1993), EPA (2/1/1993), EPA contractor updated (8/2003) contractor updated review review (8/2003) (8/2003) DEP CAS 0.8 mg/kg-d; NOAEL, 750 Not available Not available; âDâânot Not available; âDâânot classifiable 84-66-2 mg/kg-d; uncertainty factor, classifiable (2/1/1993); EPA (2/1/1993); EPA contractor updated 1,000 contractor updated review review (9/2002) Critical effects from rat (9/2002) subchronic feeding study: decreased growth rate, decreased food consumption, and altered organ weights Low confidence (2/1/1993) EPA contractor updated review (9/2002) DBP CAS 0.1 mg/kg-d; NOAEL, 125 Not available (10/1/1990); EPA Not available; âDâânot Not available; âDâânot classifiable 84-74-2 mg/kg-d; uncertainty factor, contractor updated review classifiable (2/1/1993); EPA (2/1/1993); EPA contractor updated 1,000 (11/2001) contractor updated review review (11/2001) Critical effect from rat (11/2001) subchronic-to-chronic oral study: increased mortality Low confidence (8/1/1990) EPA contractor updated review (11/2001) (Continued) 75
76 TABLE 4-1 Continued Phthalateb Chronic Oral Reference Chronic Inhalation Reference CAS No. Dose (mg/kg-d) Concentration (mg/m3) Oral Slope Factor (mg/kg-d)-1 Inhalation Unit Risk (Âµg/m3)-1 Proposed (6/2006): 0.3 mg/kg-d (acute, short term, subchronic, chronic); NOAEL, 30 mg/kg-d; uncertainty factor, 100; critical effect from rat developmental oral gavage study: developmental (decrease in fetal testosterone) BBP CAS 0.2 mg/kg-d; NOAEL, 159 Not available Not available; âCââpossible Not available; âCââpossible human 85-68-7 mg/kg-d; uncertainty factor, human carcinogen (2/1/1993); carcinogen (2/1/1993); EPA 1000 âqualitative weaknesses of the contractor updated review (8/2003) Critical effects from 6-mo rat mononuclear cell leukemia feeding study: significantly response do not provide a increased liver-to-body weight compelling basis to model the and liver-to-brain weight ratios dose-response data,â EPA contractor updated review Low confidence (2/1/1993) (8/2003) EPA contractor updated review PPRTV: 1.9 Ã 10-3 (10/1/2002); (8/2003) pancreatic cancer in male rats DEHP CAS 0.02 mg/kg-d; LOAEL 19 Not available 1.4 Ã 10-2; âB2ââprobable Not available 117-81-7 mg/kg-d; uncertainty factor, human carcinogen (2/1/1993); 1000 âorally administered DEHP Critical effect from guinea pig produced significant dose-related subchronic-to-chronic oral increases in liver tumor bioassay: increased relative liver responses in rats and mice of weight both sexesâ Medium confidence (5/1/1991)
ATSDR MRL:d 0.06; uncertainty California EPA:f Oral CSF, California EPA:f inhalation unit factor, 100; health end point, 3 Ã 10-3 risk, 2.4 Ã 10-6 reproduction (9/2002) California EPA:e 410 Âµg/d (adults); 58 Âµg/d (infant boys, age 29 days-24 months); 20 Âµg/d (neonatal infant boys, age 0-28 days) a Date of last review is shown in parentheses. Except where otherwise noted, toxicity values are from EPAâs IRIS database because these values repre- sent the highest tier in EPAâs hierarchy of toxicity values for use at Superfund sites (EPA 2003a). b EPAâs IRIS database includes a summary profile of one other phthalate: dimethyl terephthalate (DMT) (CAS 120-61-6); synonym: dimethyl-p- phthalate. However, this phthalate is not a diester of 1,2-benzenedicarboxylic acid, so it was not considered by the committee. c EPAâs Superfund Health Risk Technical Support Center developed a screening value for DMP (dated September 25, 2007) that probably falls in the third tier of the three-tier hierarchy of toxicity values. It is a subchronic RfD of 0.1 mg/kg-d that incorporates an uncertainty factor of 3,000 and is based on a LOAEL associated with increased absolute and relative liver weight and decreased serum and testicular testosterone in male rats. The authors of the PPRTV documentation concluded that âexposure to multiple phthalate esters in the environment should be taken into consideration when conduct- ing a risk assessment for DMPâ (EPA 2007b, p. 15). d Agency for Toxic Substances Disease Registry minimal risk level (ATSDR 2007a). MRLs are also available for three other phthalates to evaluate ex- posures lasting less than 1 year. e These values are Maximum Allowable Dose Levels (MADLs) for chemicals causing reproductive toxicity. Levels for male children and adolescents can be calculated by application of the default body weights in Title 22, California Code of Regulations, Section 12703(a)(8) to the procedure specified in Title 22, California Code of Regulations, Sections 12801 and 12803. California EPA also established the following MADLs for intravenous expo- sure: 4,200 Âµg/d (adults), 600 Âµg/d (infant boys, age 29 days-24 months), 210 Âµg/d (neonatal infant boys, age 0-28 days). f California EPA, Office of Environmental Health Hazard Assessment. See OEHHA (2008). 77
78 Phthalates and Cumulative Risk Assessment: The Tasks Ahead to multiple phthalate esters in the environment should be taken into considera- tion when conducting a risk assessment for DMP,â justifying the statement with the observation that âseveral phthalate esters may have a common endpoint of toxicity related to developmental and reproductive effectsâ (EPA 2007b, p. 15). The hierarchyâs entries clearly are largely out of date; any specialized risk assessment of phthalates would presumably consult the recent literature and take account of reproductive toxicity. However, at, for example, a Superfund site, multiple phthalates might be evaluated with the values in Table 4-1. Special Cases For some chemicals or chemical classesâsuch as anticholinesterase- acting pesticides, PCDDs and PCDFs, polychlorinated biphenyls (PCBs), and polycyclic aromatic hydrocarbonsâEPA has adopted special approaches that incorporate cumulative risk assessment. Those chemicals are discussed in the section âCurrent Environmental Protection Agency Cumulative Risk Assess- ment Examples and Case Studiesâ below. When chemicals or exposure circumstances are not suitably matched by any toxicity values in the defined hierarchy of sources discussed above, those performing risk assessments for EPA, such as for Superfund sites, may call on EPAâs National Center for Environmental Assessment for assistance (this would presumably occur for phthalates not included in Table 4-1). For others doing risk assessments, evaluation of toxicity values for use in risk assessments is a matter of individual choice. Some toxicity values may be derived by a risk as- sessor, for example, for assessments appearing in the peer-reviewed literature. In general, the context of the risk assessment dictates the method used to determine the toxicity values. Risk Characterization of Mixtures: Dose Addition and Independent Action As pointed out above, many risk assessments performed by using current EPA guidance evaluate simultaneous exposure to multiple chemicals (mixtures), multiple pathways of exposure, multiple routes of exposure, and multiple time- frames of exposure. Before discussing the standard approach to characterizing the risks posed by such exposures, it is helpful to discuss some general concepts of mixture evaluation. Many terms have been introduced into the literature to describe the com- bined effect on a particular end point of two or more agents acting simultane- ously in comparison with the effect of each agent acting alone. However, the terms have often been used confusingly, contradictorily, inconsistently, or incor- rectly (Berenbaum 1989). Some of the confusion and inconsistency in nomen- clature stems from attempts to evaluate the combined effects of multiple agents in terms of postulated mechanisms of action rather than in terms of the observed dose-response curves for a given effect. It is unnecessary (although not forbid-
Current Practice in Risk Assessment and Cumulative Risk Assessment 79 den) to take account of mechanisms of action in comparing joint effects of mul- tiple exposures with effects of exposures to single agents (Berenbaum 1989); all that is strictly necessary is information on the dose-response relationships for the individual components and information on the dose-response relationships for combinations of those components. That is the position adopted here and in the rest of this report. When agents in a mixture act together to produce an effect but do not en- hance or diminish each otherâs actions, the resulting mixture is defined to be noninteractive or dose-additive (or concentration-additive when the appropriate exposure measure is concentration). The prototypical noninteractive mixture is a combination of one agent with itself, and the general association between the dose-response relationship of a noninteractive mixture and the dose-response relationships of the individual components has been proved by using this proto- type (Berenbaum 1985). The committee notes that the literature can be confus- ing because of the (implicit or explicit) use of different definitions of noninterac- tion. However, it is also possible for a particular mixture to be noninteractive according to more than one definition. For example, a mixture could be dose- additive and follow the principle of independent action (see next section), which has also on occasion been used to define noninteraction in the literature. To define dose addition precisely in the general case, consider a mixture with doses dA of component A, dB of component B, dC of component C, and so on; this mixture produces level E of some specific effect. Suppose that the doses of the individual components that each acting alone produce level E of the same specific effect are DA, DB, DC..., where these values are set to infinity if that component does not produce the specific effect at any dose (the case of non- monotonic dose-response relationships is not considered here but does not pre- sent any great difficulties). The mixture is noninteractive, or is dose-additive, if and only if dA d d + B + C + = 1, (1) DA DB DC where the sum extends over all components of the mixture. Each combination of doses defines a particular mixture, and the set of all such combinations of doses that provide the same level E of effect is called the isobole for that level of ef- fect.4 Figure 4-1 gives an example of an isobole for a fixed effect of a two- component mixture and shows synergy, dose addition, and antagonism at differ- ent mixture ratios. 4 Geometrically, the isobole is a locally connected hypersurface in the space spanned by dose axes, and the noninteractive mixtures lie on the intersection of this hypersurface with the hyperplane defined by Equation 1.
80 Phthalates and Cumulative Risk Assessment: The Tasks Ahead 80 70 60 Dose of hexobarbital, mg/kg 50 40 Dose-additive mixture 30 Antagonistic mixtures 20 Synergistic mixtures 10 0 0 20 40 60 80 100 120 Dose of flurazepam, mg/kg FIGURE 4-1 Isobole for a defined electroencephalographic threshold in anesthesia for mixtures of flurazepam and hexobarbital. Any dose-additive mixture lies on the straight line; the intersection of the line and the isobole indicates the one mixture that is dose- additive for the defined threshold. The solid curve representing the isobole is an ad hoc interpolation between the measured points. Source: Norberg and WahlstrÃ¶m 1988, Table 1. Reprinted with permission; copyright 1988, Archives Internationales De Pharmacody- namie. The statement defining dose addition says nothing about the shapes of the dose-response curves for the individual components, and nothing can be ad- duced about the dose additivity or non-dose additivity of a mixture from the shapes of the dose-response curves of its components (undocumented statements to the contrary in EPA 2000, Section 4.2.2 and Table B-1, notwithstanding). Several observations about the definition of dose addition are of particular inter- est. â First, the dose additivity of a particular mixture does not imply the dose additivity of other mixtures of the same components. Mixtures of the same com- ponents may be non-dose additive for different component dosesâindeed, may be synergistic (producing an effect larger than expected for dose addition, with the sum of the left side of Equation 1 smaller than unity) at some combinations of component doses, and antagonistic (producing an effect smaller than expected
Current Practice in Risk Assessment and Cumulative Risk Assessment 81 for dose addition, with the sum of the left side of Equation 1 larger than unity) at others. Figure 4-1 provides a striking example of such a situation for a two- component mixture. â Second, conclusions about dose addition, synergism, or antagonism or more generally about the shape of an isobole may not be the same for different levels of effect even for similar mixture ratiosâgeometrically, isoboles are not necessarily parallel. â Third, the doses DA, DB, DC., ... vary with the effect level; indeed, they are just the inverses of the individual dose-response curves of the components. Box 4-2 shows how to derive the dose-additive multiple-dose-response curve for the mixture from the individual dose-response curves of the components. â Fourth, with the definition of dose addition stipulated by Equation 1, the evaluation of dose-additivity or nonadditivity is a matter entirely for obser- vation using measured dose-response curves; no consideration of mechanism of action is required. â Fifth, it is not necessary to determine any dose-response curve fully to evaluate additivity or nonadditivity of a mixture at a specific level of effect E. What is required are the doses of the individual components that, acting alone, would give a response level E and the component doses of the mixture that would give a response level E. Then if Equation 1 holds, that mixture is dose- additive at response level E. Independent Action or Response Addition An alternative approach to the comparison of the effect of a multicompo- nent mixture with the effects of individual components is what is often called independent action (also referred to as response addition or Bliss addition). The approach is based on analysis of mechanisms that depend on probabilistically independent events. If P(A) and P(B) are the probabilities for independent events A and B, respectively, to occur, and Q(A) = 1 â P(A) and Q(B) = 1 â P(B) are the corresponding probabilities for events A and B to not occur, then the probability of occurrence of event A or B (that is, occurrence of event A, or event B, or both events) is given by Equation 2, and the probability of nonoccur- rence of events A and B is given by Equation 3. P ( A âª B) = P ( A ) + P ( B) â P ( A ) P ( B) . (2) Q ( A âª B) = Q ( A ) Q ( B) . (3) Making the âconceptual leap of substituting fractional effect Eâ of an agent âfor probability of occurrence of an event, and the fractional lack of effect
82 Phthalates and Cumulative Risk Assessment: The Tasks Ahead BOX 4-2 Derivation of the Dose-Additive Multiple- Dose-Response Relationship If the individual dose-response curves for the specific effect consid- ered are fA(d), fB(d), fC(d), ... , respectively, then f A ( DA ) = E f B ( DB ) = E fC ( DC ) = E â¦ or DA = f A â1 (E) DB = f B â1 (E) DC = fC â1 (E) â¦ at effect level E (the second line uses the notation f â1 for the inverse func- tion), and there is no requirement for these doses to have the same relative values at different effect levels. Thus Equation 1 may be rewritten as dA dB dC + + + â¦ = 1. f A â1 (E) f B â1 (E) f C â1 (E) This equation is an implicit dose-additive multiple-dose-response rela- tionship for the mixture in terms of effect level E, the individual doses (dA, dB, dC, ...) of the components, and the individual dose-response relation- ships (fA, fB, fC, â¦) of the components. (i.e. fractional survival S) for probability of non-occurrenceâ (Berenbaum 1989) leads to the hypothesis of additivity of effect, or multiplication for survival in the form Eab ( d a , db ) = Ea (d a ) + Eb (db ) â Ea (d a ) Eb (db ) and (4) S ab ( d a , db ) = S a (d a ) Sb (db ), where Ea(da) now represents the fractional effect of agent a at dose da and simi- larly for agent b, and Eab and Sab are the fractional effect and its complement for the mixture of agents a and b at doses da and db. Some authors use the term ef- fect addition when the negative term in the first of Equations 4 is omitted; this is clearly inadequate theoretically for large effects (because it may lead to frac- tional effects larger than unity), but it is quite adequate for its typical use of combining small effects where the product term is negligible. Equations 4 and their generalizations to multiple agents define independ- ent action in the same way that Equation 1 defines dose addition. There is no necessary contradiction between the relationships as defined; one, both, or nei-
Current Practice in Risk Assessment and Cumulative Risk Assessment 83 ther may apply to any particular mixture of agents. Contradictions do occur, however, if (as has occurred in the literature; see Berenbaum, 1989 for an exten- sive review) independent action and noninteraction are assumed to be equiva- lent, or when synergism and antagonism are defined by deviations from one and inappropriately compared with identical terms defined by deviations from the other. Empirical Observation vs Mechanistic Inference With the assumption of dose addition or independent action, the dose- response relationship of a mixture of components may be calculated on the basis of dose-response relationships observed for mixture components. Quite often, the predictions are similar; in other cases, they differ substantially. Where dif- ferences arise, they arise from the differences in the mathematical structure of the two models. Neither prediction is guaranteed to correspond to observations. Figure 4-2 provides a hypothetical example in which observations would fall between the predictions of dose addition and independent action; the mixture exhibits synergism with respect to independent action but antagonism with re- spect to dose addition. Box 4-3 provides the details associated with the hypo- thetical example represented in Figure 4-2. 1 0.9 0.8 0.7 Fraction of normal . 0.6 Independent action 0.5 0.4 Actual 0.3 0.2 0.1 Dose addition 0 0 0.2 0.4 0.6 0.8 1 External dose FIGURE 4-2 Hypothetical example of two-component mixture dose-response curve (see Box 4-3). The components are assumed to have the same pharmacokinetics and mecha- nism of action but different Michaelis-Menten elimination constants and mass-action binding affinity to the same receptor. The fraction of normal response is assumed to be an exponential function of minus the square of the bound fraction of receptors.
84 Phthalates and Cumulative Risk Assessment: The Tasks Ahead BOX 4-3 Hypothetical Example Represented in Figure 4-2 Consider a chemical applied at dose rate d to an organism that has an elimination rate for that chemical that is of Michaelis-Menten form. The concentration C in some target tissue at steady state will then satisfy an equation of the form Î»C d = , k +C where Î» is the maximum elimination rate (so consider dose rates less than Î»), and k is the concentration at which the elimination rate is half the maxi- mum. Now suppose that the chemical binds to a receptor according to the law of mass action, so that the fraction of bound receptors is C f = C+Z for some concentration Z, and the binding to the receptor induces a devia- tion of some response from normal with magnitude depending on the square of the bound fraction (f), so that the response size is of the form (where the normal response is R = 1) R = exp ( âhf 2 ) for some constant h. Now consider two such chemicals with the constants Î», k, Z, and h given by Parameter Chemical 1 Chemical 2 Î» 1.1 1.4 k 2 0.05 Z 1 0.5 h 75 75 The curves of Figure 4-2 are obtained from a 1:7 mixture of chemical 1 and chemical 2 where chemical 2 and chemical 1 are exact alternative receptor ligands with an effect in a mixture given by C1 + 2C2 f = . 1 + C1 + 2C2
Current Practice in Risk Assessment and Cumulative Risk Assessment 85 The committee emphasizes that dose addition does not imply toxicologic similarity (as defined by EPA 2000), nor does toxicologic similarity imply dose addition, as claimed by EPA (2000); Figure 4-2 is a hypothetical counterexam- ple of the last proposition that shows that dose addition need not apply even to mixtures of components with identical mechanisms of action. Similarly, inde- pendent action does not imply, nor is it implied by, different mechanisms for the mixture components. Nor are dose addition and independent action mutually exclusive (see Berenbaum 1989 for counterexamples of these propositions). Practical Applications, Relative Potencies, and Toxicity Equivalence Factors Evaluation of any particular mixture of agents requires an empirical de- termination of how they combine to produce any particular effect. It is generally convenient in performing such evaluations to compare observations against dose addition or independent action, and if the deviations are small enough to be sta- tistically insignificant the mixtures may be considered to exhibit dose addition or independent action for that particular effect at that effect level (many exam- ples of both kinds are known; see Berenbaum 1989 for an extensive review). As pointed out above, dose addition may apply at particular levels of effect (and the same is true of independent action), so a complete evaluation requires examina- tion of all effect levels or at least over the range of effect levels that are of prac- tical importance (effects that might be expected, given the levels of exposure under evaluation). It must be emphasized that any particular finding applies strictly only to the particular effect evaluated; for example, there is nothing to prevent one end point from exhibiting dose addition and another from departing substantially from it at the same dose. For particular types of mixtures, some plausible assumptions are generally made. Some groups of chemicals may have similar chemical structures and act and be acted on in similar ways in the body. For example, all may be absorbed in the same way (although with quantitative differences in the rate and extent of absorption), and all may be detoxified in the same organ by the same enzyme systems (although with differing Vmax and km for the rate of detoxification). In such circumstances, it is possible and may be plausible to proposeâsubject to experimental confirmationâthat each acts as the same dilution of the other at all doses and in all mixtures. Then, dose addition would apply, and each could be compared directly with some reference agent in the group by using a relative potency that specifies the effective dilution; however, the assumptions, such as similar absorption or detoxification, are not necessarily correct, nor is it neces- sary that they be satisfied for dose addition to apply. A dose of one agent is then equivalent to a multiple of the dose of another, with the multiple being the same
86 Phthalates and Cumulative Risk Assessment: The Tasks Ahead for every dose, so such agents would necessarily have parallel dose-response curves.5 Relative potency may, however, differ for different end points. Relative- potency estimates of this nature are used (or are proposed for use) for cancer potency estimates for some chemicals and for noncancer end points for some chemicals, for example, for anticholinesterase agents (see the section âCurrent EPA Cumulative Risk Assessment Examples and Case Studiesâ below). Al- though parallel dose-response curves are necessary for a relative-potency ap- proach to be correctly used, they are not sufficient, because the parallelism of dose-response curves gives no information on the effect of a mixture. An even more restrictive application of dose addition may be proposed and used in practice, as it has been for the 2,3,7,8-chlorinated dibenzo-p-dioxins and -furans and for âdioxin-likeâ PCB congeners (Van den Berg et al. 2006). For those congeners, a potency estimate, the toxicity equivalence factor (TEF), has been estimated relative to the prototype of the 2,3,7,8-tetrachlorodibenzo-p- dioxin group, but this potency is supposed to apply to every end point affected via the aryl hydrocarbon receptor (at least for âdioxin-likeâ effects in the case of the PCB congeners). By contrast, the less-restrictive relative potencies discussed above may in principle differ for different end points. Risk Characterization Practical Risk Characterization for Cancer End Points For cancer end points, the exposure estimates required to be obtained as described above under âExposure Assessmentâ for each receptor evaluated are lifetime average dose rates (expressed in units of milligrams per kilogram per day) or lifetime average exposures (expressed as air or water concentrations). For all pathways of exposure, routes of exposure, and chemicals, individual- chemical risk estimates are obtained by multiplying lifetime average dose rates by the relevant route-specific CSF or multiplying lifetime average exposure by the relevant route-specific unit risks. For brevity, âthresholdâ carcinogens, for which reference doses and reference concentrations are used in the same manner as noncancer end points, are not addressed here (see EPA 2005a for further guidance). In special circumstances, such as for exposures primarily of young children (EPA 2005b), the standard CSFs may be modified to take account of increased susceptibility of the exposed population. 5 To be consistent with EPA nomenclature, parallel dose-response curves mean that âfor equal effects, the dose of one component is a constant [positive] multiple of the dose of a second componentâ (EPA 2000). The committee added the term positive for preci- sion because dose-response curves that are related by a negative multiplier are not con- sidered parallel.
Current Practice in Risk Assessment and Cumulative Risk Assessment 87 A total cancer risk estimate is then obtained by summing the individual- chemical risk estimates obtained across pathways, routes of exposure, and chemicals. The first two summations, of pathways and routes, are for a single chemical and thus involve no assumptions about interactions between chemicals. The last summation, of chemicals, explicitly incorporates an assumption of in- dependent action (strictly, effect addition). For the end point of interest, lifetime probability of cancer, all carcinogens are treated as having exactly additive ef- fects. For carcinogens at low doses (that is, at lifetime average dose rates low enough for the predicted probability of cancer to be substantially smaller than 0.1âand in practice almost always smaller than 0.001), it is also assumed that the dose-response curve is linear with no threshold. Thus, the effect-addition assumption is in this case also a dose-additive assumption. Because estimated responses are generally sufficiently small for the negative product terms to be negligible in the generalization of Equation 4 to multiple chemicals, independent action and effect addition are equivalent to the accuracies required. Cases in which estimated dose rates are high enough for this not to be true would be treated as obvious emergencies for which any alternative treatment would be unnecessary. Practical Risk Characterization for Noncancer End Points For noncancer end points, the exposure estimates required to be obtained as discussed in the section âExposure Assessmentâ for each receptor evaluated are average dose rates (expressed in units of milligrams per kilogram per day) or exposures (expressed as air or water concentrations) over a lifetime and over various shorter periods (at any age, but particularly in childhood). For each chemical, each pathway, and each averaging period, a hazard quotient (HQ) is calculated as Average dose rate Average concentration HQ = â routes RfD or RfC , (5) where the summation is over all routes of exposure, and the RfD or RfC used is appropriate to the averaging period and route or has been adjusted from an alter- native averaging period or route to be appropriate. If shorter-term RfD or RfC values (or equivalents from the hierarchic selection of toxicity values) are not available, RfD and RfC values from longer-term exposures may be used for shorter-term exposures, and the resulting HQs are considered likely to be con- servative (overestimates). For each averaging period, an overall summary hazard index (HI) is then calculated as the sum of HQs for each pathway and each chemical, so
88 Phthalates and Cumulative Risk Assessment: The Tasks Ahead HI = â â pathways chemicals HQ. (6) The summation over chemicals here is explicitly chosen to be a special case of the summation used in the definition of dose addition (compare Equation 1), in which the RfC or RfD used in Equation 5 (and hence in Equation 6) corre- sponds to a dose rate that has the same effect (namely, no effect) for each chemical. Thus, under the hypothesis of dose addition, if the HI is less than or equal to unity, no effect can be expected from the mixture of chemicals incorpo- rated in the summation. However, the RfC or RfD is not necessarily the largest dose rate or concentration that would result in no effect, so HIs larger than unity cannot necessarily be taken to indicate a larger than zero effect of the mixture under the dose-addition hypothesis, although they are treated as indicators that there is potentially such a nonzero effect. Thus, if the summary HI is less than or equal to unity, there is unlikely to be an appreciable risk of deleterious effects, and the analysis is usually com- plete. If the summary HI is larger than unity, further analysis may be performed that takes account of âeffectâ and âmechanism of actionâ in an attempt to deter- mine whether application of dose addition to all the chemicals simultaneously is justifiable or to determine whether the RfD or RfC used in evaluating the sum- mation is appropriate for any particular common effect or mechanism (see the section âEmpirical Observation vs Mechanistic Inferenceâ above). EPA guidance on segregation of chemicals by effect is shown in Box 4-4. The guidance states further that âif one of the effect-specific hazard indices ex- ceeds unity, consideration of the mechanism of action might be warranted. A strong case is required, however, to indicate that two compounds which produce adverse effects on the same organ system (e.g., liver), although by different mechanisms, should not be treated as dose additive. Any such determination should be reviewed by ECAO [Environmental Criteria and Assessment Office in EPAâs Office of Health and Environmental Assessment]â (EPA 1989a, p. 8-14). It is further pointed out that obtaining the information required to segre- gate chemicals by effect or mechanism of action is difficult to locate (see Box 4- 5 below). Furthermore, âif there are specific data germane to the assumption of dose-additivity (e.g., if two compounds are present at the same site and it is known that the combination is five times more toxic than the sum of toxicities for the two compounds), then modify the development of the hazard index ac- cordingly. Refer to the EPA (1986) mixtures guidelines for discussion of a haz- ard index equation that incorporates quantitative interaction data. If data on chemical interactions are available, but are not adequate to support a quantitative assessment, note the information in the âassumptionsâ being documented for the site risk assessmentâ (EPA 1989a, pp. 8-14).
Current Practice in Risk Assessment and Cumulative Risk Assessment 89 BOX 4-4 Procedure for Segregation of Hazard Indexes by Effect Segregation of hazard indices requires identification of the major ef- fects of each chemical, including those seen at higher doses than the criti- cal effect (e.g., the chemical may cause liver damage at a dose of 100 mg/kg-day and neurotoxicity at a dose of 250 mg/kg-day). Major effect categories include neurotoxicity, developmental toxicity, reproductive toxic- ity, immunotoxicity, and adverse effects by target organ (i.e., hepatic, renal, respiratory, cardiovascular, gastrointestinal, hematological, musculoskele- tal, and dermal/ocular effects). Although higher exposure levels may be required to produce adverse health effects other than the critical effect, the RfD can be used as the toxicity value for each effect category as a conser- vative and simplifying step. Source: EPA 1989a. BOX 4-5 Information Sources for Segregation of Hazard Indexes Of the available information sources, the ATSDR Toxicological Pro- files are well suited in format and content to allow a rapid determination of additional health effects that may occur at exposure levels higher than those that produce the critical effect. Readers should be aware that the ATSDR definitions of exposure durations are somewhat different than EPAâs and are independent of species; acuteâup to 14 days; intermedi- ateâmore than 14 days to 1 year; chronicâgreater than 1 year. IRIS con- tains only limited information on health effects beyond the critical effect, and EPA criteria documents and HEAs, HEEPs, and HEEDs may not systemati- cally cover all health effects observed at doses higher [than] those associ- ated with the most sensitive effects. Source: EPA 1989a. Special Considerations in Practical Risk Characterizations As pointed out above, some groups of chemicals are treated specially by using relative-potency or TEF approaches; these are discussed further in the section âCurrent EPA Cumulative Risk Assessment Examples and Case Stud- iesâ below. Some mixtures, such as Aroclors (PCB mixtures), may be treated as individual chemicals in toxicity assessments because they have been tested in toxicity studies. However, it is unlikely that the precise mixtures tested (and there may be some doubt as to their characterization in any case) will ever be
90 Phthalates and Cumulative Risk Assessment: The Tasks Ahead what receptors are exposed to after transport through the environment, so actual exposures are likely to be to mixtures with congener or other component profiles differing from those tested. There are also situations in which the risk assess- ments required do not correspond completely to the âtypicalâ assessment de- scribed here, such as nationwide evaluations of cumulative and aggregate expo- sures to pesticides (cumulative refers to the multiple-chemical nature of the assessment and aggregate to the multiple pathways of exposure). Summary of Current Risk-Assessment Approaches In summary, the usual approach to EPA-style risk assessments for non- cancer end points is initially âdose-additiveâ for all chemicals, partly to ensure an initial conservative assessment. Later, if such a conservative approach does not suffice, the dose-addition approach is applied independently to subsets of chemicals with the same end point or mechanism, where mechanism is not well defined. For cancer end points, the usual approach of summing risk estimates for all chemicals is both response-additive and dose-additive because the two are equivalent when the standard low-dose linear hypothesis is used.6 In every case, direct information on any particular mixture that contradicts the hypothesis of dose addition will override the default approach. THE EVOLUTION OF GUIDANCE ON CUMULATIVE RISK ASSESSMENT Table 4-2 summarizes the evolution of EPA guidance (or, for the Interna- tional Life Sciences Institute document, in cooperation with EPA) on cumulative risk assessment. Undoubtedly, other documents have influenced the practice of cumulative risk assessment, but the committee believes that those cited here have been the primary sources for EPA consideration of cumulative risk assess- ment. Table 4-2 summarizes the stated purpose of the guidance, the definitions of cumulative adopted in the guidance document, and the default approach taken for evaluation of cumulative risks posed by mixtures of chemicals and other stressors when there is no direct information on the particular (or sufficiently similar) mixtures (so that the effect of the mixture has to be estimated from measured effects of individual components). As far as possible, the committee has quoted the documents or relevant memoranda accompanying the documents on their release for the summaries. At times, that proved difficult because there may be more than one statement or definition, and the default approach may not have been explicitly stated. The âDefault approachâ column of the table high- lights some statements made in the guidance about the conditions required for dose addition or independent action. 6 For brevity, âthresholdâ carcinogens were not addressed here.
TABLE 4-2 Summary of Stated Purposes of Guidance Documents, Definition of Cumulative in the Context of Risk Assessment, and the Default Approach Taken to the Toxicity of Mixtures of Chemicals Document Title, Author Agency, Date Stated Purpose Definition of Cumulative Default Approach Guidelines for the Health Risk âGenerate a consistent Agency approach Not defined Although all authorities cited adopted dose- Assessment of Chemical for evaluating data on the chronic and additive models, âdose additive models are Mixtures, EPA/630/R-98/002, subchronic effects of chemical mixturesâ not the most biologically plausible approach September 1986, Risk (EPA 1986, p. 1) if the compounds do not have the same mode Assessment Forum, also of toxicologic action,â and the published 51FR34014-34025 recommendation was that âdepending on the nature of the risk assessment and the available information on modes of action and patterns of joint action, the â¦ most reasonable additive model should be used.â (Dose addition and independent action are the only alternatives discussed.) Later, however, a hazard-index approach (dose addition) is recommended for systemic toxicants, although âsince the assumption of dose addition is most properly applied to compounds that induce the same effect by similar modes of action, a separate hazard index should be generated for each end point of concern.â (EPA 1986, pp. 8-9) Technical Support Document Supplement to 1986 guidelines Limited to mixtures of chemicals None. on Risk Assessment of Chemical Mixtures, EPA/600/8-90/064, November 1988, Office of Research and Development (Continued) 91
92 TABLE 4-2 Continued Document Title, Author Agency, Date Stated Purpose Definition of Cumulative Default Approach Risk Assessment Guidance for âDeveloped to be used in the remedial Not defined; cites to 1986 guidelines Default is dose addition; hazard quotients are Superfund, Volume 1, Human investigation/feasibility study (RI/FS) for using dose addition for summed to produce a hazard index (HI). âIf Health Evaluation Manual process at Superfund sites, although the âaggregateâ risks of multiple the HI is greater than unityâ¦it would be (Part A), Interim Final, analytical framework and specific chemicals; âalthough the calculation appropriate to segregate the compounds by EPA/540/1-89/002, December methods described in the manuals may procedures differ for carcinogenic effect and by mechanism of action.â 1989 also be applicable to other assessments and noncarcinogenic effects, both Furthermore, âif one of the effect-specific of hazardous wastes and hazardous sets of procedures assume dose hazard indices exceeds unity, consideration of materialsâ (EPA 1989a, p. xv) additivity in the absence of the mechanism of action might be warranted. information on specific mixturesâ A strong case is required, however, to (EPA 1989a, pp. 8-12) Chemicals indicate that two compounds which produce and radiation only adverse effects on the same organ system (e.g., liver), although by different mechanisms, should not be treated as dose additive.â (EPA 1989a, p. 8-14) Guidance on Cumulative Risk âThis guidance directs each office to take âAdverse health and ecological âDue to the current state of the practice and Assessment, Part I Planning into account cumulative risk issues in effects from synthetic chemicals, limited data, the aggregation of risks may and Scoping, EPA Science scoping and planning major risk radiation, and biological stressors,â often be based on a default assumption of Policy Council, July 3, 1997; assessments and to consider a broader âsocial, economic, behavioral or additivityâ (EPA 1997b, p. 3) (there is no memo from EPA scope that integrates multiple sources, psychological stressors that may definition of additivity.) administrator, July 3, 1997, effects, pathways, stressors and contribute to adverse health effects,â quoted populations for cumulative risk analyses including âexisting health condition, in all cases for which relevant data are anxiety, nutritional status, crime, available.â (EPA 1997a, p. 1) and congestion.â (EPA 1997b, p. 2) A Framework for Cumulative âThe goal of this project is to develop a Implicitly examined only multiple No default is specified, although the text may Risk Assessment, ILSI Risk framework that can be used to guide the chemical exposures imply some sort of unspecified additivity at Science Institute Workshop conduct of cumulative risk assessments.â low dose. Report (1999) (ILSI 1999, p. 2)
Supplementary Guidance for âThis document describes more detailed Examines only chemical mixtures, Default of hazard index or relative potency Conducting Health Risk procedures for chemical mixture intended as only a component of a for âtoxicologically similarâ components, Assessment of Chemical assessment using data on the mixture of cumulative assessment as described independent action for âtoxicologically Mixtures, EPA/630/R-00/002, concern, data on a toxicologically similar by EPA 1997b (above) independentâ components, interaction hazard August 2000; Risk Assessment mixture, and data on the mixture index for âinteractionsâ Forum component chemicals.â (EPA 2000, âIn practice, because of the common lack of p. xiv) information on mode of action and pharmacokinetics, the requirement of toxicologic similarity is usually relaxed to that of similarity of target organs (U.S. EPA, 1989a).â (EPA 2000, p. 80) Citation is to guidance listed above. Guidance on Cumulative Risk âThis document provides guidance to âThis document provides guidance âA cumulative risk assessment begins with Assessment of Pesticide OPP [Office of Pesticide Programs] to OPP scientists for evaluating and the identification of a group of chemicals, a Chemicals That Have a scientists for evaluating and estimating estimating the potential human risks Common Mechanism Group (CMG), that Common Mechanism of the potential human risks associated with associated with such multichemical induce a common toxic effect by a common Toxicity, EPA Office of such multichemical and multipathway and multipathway exposures to mechanism of toxicity.â (EPA 2002, p. iii) Pesticide Programs, January exposures to pesticides.â (EPA 2002, pesticides. This process is referred EPA here assumes that identification of a 14, 2002 p. ii) to as cumulative risk assessment.â CMG implies dose addition for its member (EPA 2002, p. ii) chemicals, so dose addition is the only possibility considered. It is also asserted (incorrectly) that âdose addition requires a constant proportionality among the effectiveness of the chemicals.â (EPA 2002, p. 31) (Continued) 93
94 TABLE 4-2 Continued Document Title, Author Agency, Date Stated Purpose Definition of Cumulative Default Approach Framework for Cumulative âImmediately offers a basic structure and âCumulative Risk: The combined No explicit default; refers to previous Risk Assessment, EPA/630/P- provides starting principles for EPAâs risks from aggregate exposures to documents, particularly EPA (2000), in which 02/001F, May 2003, Risk cumulative risk assessmentsâ (EPA multiple agents or stressors.â (EPA defaults are specified. Assessment Forum 2003b, p. 5) 2003b, p. 6) âOffers the basic principles around which âAggregate exposure: The combined to organize a more definitive set of exposure of an individual (or cumulative risk assessment guidanceâ defined population) to a specific (EPA 2003b, p. 5) agent or stressor via relevant routes, pathways, and sources.â (EPA 2003b, p. 7) Concepts, Methods and Data âThis current report serves as a resource âThe Framework defines cumulative No explicit default; refers to previous Sources for Cumulative Health document for identifying specific risk as the combined risks from documents, particularly EPA (2000), in which Risk Assessment of Multiple elements of and approaches for aggregate exposures (i.e., multiple defaults are specified. Chemicals, Exposures and implementing cumulative risk route exposures) to multiple agents Effects: A Resource assessments. This report is not a or stressors, where agents or Document. EPA/600/R- regulatory document and is not guidance stressors may include chemicals, as 06/013F, August 2007, EPA, but rather a presentation of concepts, well as biological or physical agents National Center for methods and data sources.â (EPA 2007g, (e.g., noise, nutritional status), or the Environmental Assessment, p. xvii) absence of a necessity such as Cincinnati, OH habitat (U.S. EPA, 2003a). Cumulative risk assessment, then, is an analysis, characterization and possible quantification of the combined risks to health or the environment from multiple agents or stressors.â (EPA 2007g, p. xxi), citation in the quote is to the framework listed above.
Current Practice in Risk Assessment and Cumulative Risk Assessment 95 CURRENT ENVIROMENTAL PROTECTION AGENCY EXAMPLES AND CASE STUDIES OF CUMULATIVE RISK ASSESSMENT Cumulative risk assessment is not new, although development and appli- cation of relevant EPA guidance continues to evolve (see, for example, EPA 2007g). EPAâs IRIS database includes toxicity values for chemical mixtures, such as coke-oven emissions, diesel-engine exhaust, PCBs, xylene isomers, a 2,4- and 2,6-dinitrotoluene mixture, and a 2,4- and 2,6-toluene diisocyanate mixture. In addition, Table 4-3 highlights recent applications of cumulative risk assessment to evaluate human exposure to chemicals. The following sections provide more detailed descriptions of two EPA programs that involve cumula- tive evaluations of pesticides and air toxics. Aggregate and Cumulative Assessments of Pesticides EPAâs Office of Pesticide Programs implements a two-stage assessment process for groups of pesticides that have a common mechanism of toxicity. First, an aggregate assessment that considers all pathways and routes of expo- sure of each member of the group is completed (EPA 2008d,e); depending on the results, risk-reduction actions may be taken. Then a cumulative assessment considers exposure of and risks to all members of the group; additional risk- reduction steps may be taken on the basis of the results. Risk-reduction actions include elimination or restriction of pesticide uses. Cumulative risk assessments of pesticides with a common mechanism of toxicity involve extensive dose-response modeling for each pesticide, which provides the relative potencies used in the dose-additivity-based cumulative method for common-mechanism pesticides (EPA 2002). Such risk assessments also involve a multicomponent exposure assessment (EPA 2002). Dietary expo- sures are estimated from nationally representative dietary and pesticide-residue surveys. Drinking-water exposures and residential and nonoccupational pesti- cide uses are estimated by region to reflect variations in agriculture, pest pres- sures, and home and other pesticide uses. The datasets are compiled into an in- dividual-level daily-exposure estimate over the course of a year. For the risk characterization, relevant durations of exposure are defined, and rolling-average exposures to individuals are developed on the basis of the daily-exposure esti- mates (EPA 2002). As implied in the descriptions of dose-response and expo- sure-assessment procedures, cumulative risk assessments of common- mechanism pesticides involve consideration of the timing and duration of expo- sures and the timing of onset and duration of health effects and recovery (EPA 2002).
96 TABLE 4-3 Summary of Cumulative Human Risk Assessment Applications to Evaluation of Chemical Exposuresa Chemical Mixture Cumulative Risk Assessment Approach References Asbestos fibers Asbestos includes various naturally occurring silicate fibers, and their cancer potency may vary as a function of fiber EPA 2008f type and size. Therefore, EPA developed draft guidance, currently undergoing review by its Science Advisory Board, that provides an approach for quantifying differences in cancer potency among fiber types (amphibole or chrysotile) and particle sizes (length and width). Carcinogenic EPA classifies benzo[a]pyrene (B[a]P) and six other polycyclic aromatic hydrocarbons (PAHs) as B2 carcinogens. EPA 1993b; polycyclic Results are consistent among cancer bioassays involving B[a]P and these PAHs; however, insufficient data are Carlson-Lynch aromatic available to derive cancer slope factors for all these PAHs. Also, although these PAHs may cause cancer by the same et al. 2007 hydrocarbons mechanism as B[a]P, they appear to be less potent. EPA developed a relative-potency approach to estimate cancer risk associated with these PAHs by comparing PAH cancer potencies, using skin tumorogenicity bioassays, and quantifying âorder of magnitudeâ relative potency factors (RPFs) for the six carcinogenic PAHs on the basis of comparison with the index chemical, B[a]P. This RPF approach can be used to evaluate PAH mixtures as they occur in the environment, with proportions depending on source, age of release, and environmental conditions. EPA is re- evaluating the toxicity of B[a]P and recently presented preliminary analyses in which EPA defined 26 PAHs, instead of the current six, with adequate data for RPF derivation. Dioxin-like People are exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD, or âdioxinâ) and other 2,3,7,8-chlorinated EPA1987, 1989b; chemicals dioxin congeners and dioxin-like compounds as complex mixtures. Seven dioxin, 10 furan, and 12 polychlorinated Van den Berg et al. biphenyl (PCB) congeners may exert toxic effects through the same mechanism of action as 2,3,7,8-TCDD, namely, 1998, 2006 binding to the aryl hydrocarbon receptor (AhR), a cellular protein. A toxic equivalence (TEQ) approach has been developed to estimate risk associated with 2,3,7,8-TCDD and other dioxin-like congeners. The approach applies to AhR-mediated effects, assuming a model of dose addition. Each dioxin-like congener has been assigned a toxic equivalence factor (TEF) to represent the fractional toxicity of the congener relative to that of 2,3,7,8-TCDD. TEFs are used to transform concentrations of individual dioxin-like congeners into an equivalent concentration of 2,3,7,8- TCDD, as determined by the equation TEQ = â [( Ci )( TEFi )] + â [( Ci )( TEFi )] + â [( Ci )( TEFi )] , iâPCDDs iâPCDFs iâPCBs
where TEQ = equivalent TCDD concentration, TEFi = toxic equivalency factor for congener i, and Ci = concentration of congener i. This TEQ estimate is combined with toxicity data on 2,3,7,8-TCDD to quantify risk posed by exposure to dioxin-like congener mixtures. Polychlorinated Commercial PCB mixtures released into the environment may be altered as a result of environmental processes, such EPA 1996a,b,c; biphenyls as partitioning, transformation, and bioaccumulation through the food chain. Therefore, EPA recommends an approach EPA 1997c to assess cancer risk associated with exposure to PCBs that accounts for different PCB mixtures typically found in environmental media. Cancer studies to date suggest that more highly chlorinated, less volatile congeners are associated with greater cancer risk. Those congeners tend to persist in the environment in soil and sediment and to bioaccumulate in biota. More volatile, less chlorinated congeners that partition into air or surface water are more likely to be metabolized and eliminated than highly chlorinated congeners. Therefore, EPA recommends using the environmental medium or exposure medium as an indicator of the cancer potency of a PCB mixture. For noncancer effects, EPA has developed reference doses for two commercial PCB mixtures (Aroclor 1016 and Aroclor 1254), which account for the toxicity of the mixtures but not necessarily how they might have changed after release into the environment. Petroleum The composition of petroleum products changes after release into the environment. For that reason, use of toxicity data MADEP 2002, hydrocarbon on whole products may be appropriate for fresh spills but not for older spills that have had time to weather. 2003; Edwards et fractions Alternatively, evaluating only a subset of individual chemicals in a mixture, such as carcinogenic PAHs and benzene, al. 1997; might not account for toxicity associated with the rest of the mixture. Therefore, a fraction-based approach was devised Gustafson et al. that consists of dividing petroleum mixtures into fractions and assigning physical and chemical properties and toxicity 1997 values to each fraction. This approach accounts for environmental weathering of spilled product and is a practical alternative to evaluation of hundreds of individual petroleum chemicals. Furthermore, data on toxicity and fate and transport properties needed for assessing health risk are not available for many petroleum hydrocarbons. ATSDR The Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) directs ATSDR, where ATSDR 2004, interaction feasible, to develop methods for determining the health effects of substances in combination with other substances with 2007b profiles which they are commonly found. Exposure to two or more chemicals is common at hazardous-waste sites that ATSDR evaluates. Therefore, ATSDR developed a chemical-mixtures program in response to the CERCLA directive, including identification of mixtures of highest concern for pubic health and publication of final interaction profiles that evaluate toxicity data on a whole mixture, where available, and otherwise rely on mostly binary data relevant to the (Continued) 97
98 TABLE 4-3 Continued Chemical Mixture Cumulative Risk Assessment Approach References joint toxic action of chemicals in the mixture. ATSDR has completed profiles for the following mixtures: (1) arsenic, cadmium, chromium, and lead; (2) benzene, toluene, ethylbenzene, and xylenes; (3) lead, manganese, zinc, and copper; (4) persistent chemicals found in breast milk; (5) persistent chemicals found in fish; (6) 1,1,1-trichloroethane, 1,1-dichloroethane, trichloroethylene, and tetrachloroethylene; (7) cesium, cobalt, PCBs, strontium, and trichloroethylene; (8) arsenic, hydrazines, jet fuels, strontium-90, and trichloroethylene; (9) cyanide, fluoride, nitrate, uranium; (10) atrazine, deethylatrazine, diazinon, nitrate, simazine; and (11) chlorpyrifos, lead, mercury, and methylmercury. Disinfection People are exposed simultaneously to disinfection byproducts in drinking water. EPA developed an approach to EPA 2003c; byproducts cumulative risk assessment of disinfection-byproduct mixtures that requires exposure modeling and physiologically Teuschler et al. based pharmacokinetic modeling combined with the use of cumulative relative potency factors (CRPFs). The use of 2004 CRPFs provides multiple-route, chemical-mixture risk estimates based on total absorbed doses. a Pesticides and air toxics are described in detail in text.
Current Practice in Risk Assessment and Cumulative Risk Assessment 99 No new regulatory actions were needed on the basis of EPAâs recent cu- mulative assessment of 10 N-methyl carbamate pesticides because actions taken on the basis of aggregate assessments of the individual pesticides had achieved necessary risk reductions (EPA 2008g). For example, all domestic uses of carbo- furan were deemed ineligible for reregistration, given the findings of its aggre- gate assessment. All U.S. uses of carbofuran will be canceled (EPA 2007c). National Air Toxics Assessment The National Air Toxics Assessment (NATA) is a national assessment of health risks associated with inhalation of 33 hazardous air pollutants (air toxics) and diesel particulate matter (qualitative assessment only). Assessment results are disseminated online for the public and used to inform the air-toxics program in priority-setting, air-pollution trends assessment, research, and planning (EPA 2007d). The NATA estimates concurrent exposures to the selected chemicals at the census-tract, county or state level at a selected time (EPA 2006). The cumulative methods applied for the NATA are dose addition and independent action. The common noncancer health effect of concern is respiratory irritation (irritation of the lining of the respiratory system), and single-chemical HQs of respiratory irritants are added to yield a ârespiratory hazard indexâ (dose addition). For the carcinogens, lifetime cancer risk estimates for inhalation exposures are added (independent action but also in effect dose addition because of the assumed dose-response linearity) (EPA 2007e). More than 25 million people live in census tracts where air pollutants con- tribute to upper-bound estimates of more than 10 in 1 million increment in life- time cancer risk. The most important carcinogens that are known to contribute to the estimated excess risks are benzene and chromium (EPA 2007f). STRENGTHS AND WEAKNESSES OF CURRENT APPROACHES OR PRACTICES Having reviewed the current cumulative risk assessment practices and approaches, the committee has made the following observations: â EPA has been addressing cumulative impact and risk under various le- gal and regulatory authorities. â Various offices and organizations in the EPA have devoted consider- able resources to developing concepts and guidance regarding cumulative risk assessment. â In cumulative risk assessments of human health effects, there is a reli- ance on dose addition as the default approach.
100 Phthalates and Cumulative Risk Assessment: The Tasks Ahead â Current practices focus on well-defined mixtures of chemical stressors to which simultaneous (or concurrent) exposures occur. In its Framework for Cumulative Risk Assessment (EPA 2003b), EPA has developed an appropriately broad definition of cumulative risk assessment and identified multiple approaches to the conduct of such assessments. EPA, through its various offices, has accrued substantial practical experience with cumulative risk assessment. However, the assessments conducted to date have been of well- defined groups of chemicals to which simultaneous exposure occurs. Chemicals are grouped according to a common mechanism of toxic action or end point and specific exposure situations, such as a hazardous-waste site or spill or presence in food or water. Therefore, although multiple methods are available, EPA has used only a few of them in practice. And despite recognition of nonchemical stressors as potentially important contributors to cumulative risk, nonchemical stressors are rarely addressed or evaluated. APPLICATION TO PHTHALATES EPA clearly has given considerable thought to cumulative risk assessment and has produced substantial guidance on it. On the basis of that guidance, a mixture of phthalates should be included in a cumulative assessment based on âtoxicologic similarityâ (see Chapter 3). However, there may be inconsistencies in how different offices in EPA would perform risk assessments, the available IRIS toxicity values do not incorporate the relevant end points that would sug- gest toxicologic similarity, and some of the guidance is pulling in different di- rections in that toxicologic similarity is largely undefined. A sufficiently de- tailed examination of the toxicologic profiles and mechanisms of action of the individual phthalates would find distinct differences in end points affected or the degree to which specific end points are affected and in detailed mechanisms of action, so toxicologic similarity would be ambiguous. The following chapter examines the evaluation of phthalate mixtures in more detail and provides practical approaches to the examination of phthalates mixtures in particular and other mixtures in general. REFERENCES ATSDR (Agency for Toxic Substances and Disease Registry). 2004. Guidance Manual for the Assessment of Joint Toxic Action of Chemical Mixtures. U.S. Department of Health and Human Services, Public Health Service, Agency for Toxic Sub- stances and Disease Registry. May 2004 [online]. Available: http://www.atsdr. cdc.gov/interactionprofiles/ipga.html [accessed July 22, 2008]. ATSDR (Agency for Toxic Substances and Disease Registry). 2007a. Minimum Risk Levels (MRLs) for Hazardous Substances, November 2007 [online]. Available: http://www.atsdr.cdc.gov/mrls/ [accessed June 16, 2008].
Current Practice in Risk Assessment and Cumulative Risk Assessment 101 ATSDR (Agency for Toxic Substances and Disease Registry). 2007b. Final Interaction Profiles [online]. Availble: http://www.atsdr.cdc.gov/interactionprofiles/ [accessed June 16, 2008]. Berenbaum, M.C. 1985. The expected effect of a combination of agents: The general solution. J. Theor. Biol. 114(3):413-431. Berenbaum, M.C. 1989. What is synergy? Pharmacol. Rev. 41(2):93-141. Carlson-Lynch, H., J. Stickney, P. McClure, M. Gehlhaus, and L. Flowers. 2007. Pro- posed Derivation of Relative Potency Factors (RPFs) for Individual Polycyclic Aromatic Hydrocarbons and Characterization of Uncertainty. Abstract M3-E5. Presented at the Society for Risk Analysis Annual Meeting 2007-Risk 007: Agents of Analysis, December 9-12, 2007, San Antonio, TX [online]. Available: http://birenheide.com/sra/2007AM/program/singlesession.php3?sessid=M3-E [ac- cessed July 15, 2008]. Edwards, D.A., M.D. Andriot, M.A. Amoruso, A.C. Tummey, C.J. Bevan, A. Tveit, L.A. Hayes, S.H. Youngren, and D.V. Nakles. 1997. Development of Fraction Specific Reference Doses (RfDs) and Reference Concentrations (RfCs) for Total Petroleum Hydrocarbons (TPH). Total Petroleum Hydrocarbon Criteria Working Group Se- ries, Vol. 4. Amherst, MA: Amherst Scientific Publishers [online]. Available: http://www.aehs.com/publications/catalog/contents/Volume4.pdf [accessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 1986. Guidelines for the Health Risk As- sessment of Chemical Mixtures. EPA/630/R-98/002. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. September 1986 [on- line]. Available: http://www.epa.gov/ncea/raf/pdfs/chem_mix/chemmix_1986.pdf [accessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 1987. Interim Procedures for Estimating Risks Associated with Exposures to Mixtures of Chlorinated Dibenzo-p-Dioxins and -Dibenzofurans (CDDs and CDFs). EPA/625/3-87/012. Risk Assessment Fo- rum, U.S. Environmental Protection Agency, Washington, DC. March 1987. EPA (U.S. Environmental Protection Agency). 1989a. Risk Assessment Guidance for Superfund, Volume I. Human Health Evaluation Manual (Part A). Interim Final. EPA/540/1-89/002. PB90-155581. Office of Emergency and Remedial Response, U.S. Environmental Protection Agency, Washington, DC. December 1989 [online]. Available: http://rais.ornl.gov/homepage/HHEMA.pdf [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 1989b. Interim Procedures for Estimating Risks Associated with Exposures to Mixtures of Chlorinated Dibenzo-p-Dioxins and -Dibenzofurans (CDDs and CDFs) and 1989 Update. EPA/625/3-89/016. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. March 1989. EPA (U.S. Environmental Protection Agency). 1992. Guidelines for Exposure Assess- ment. EPA/600/Z-92/001. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. May 1992 [online]. Available: http://rais.ornl.gov/ homepage/GUIDELINES_EXPOSURE_ASSESSMENT.pdf [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 1993a. Use of IRIS Values in Superfund Risk Assessment. OSWER Directive 9285.7-16. Memorandum to Directors, Waste Management Division Region I, IV, VII, VIII, Director, Emergency and Remedial Response Division Region II, Directors, Hazardous Waste Management Division
102 Phthalates and Cumulative Risk Assessment: The Tasks Ahead Regions III, VI, IX, and Directors, Hazardous Waste Division, Region X, from William H. Farland, Director, Office of Health and Environmental Assessment, Henry L. Longest II, Director, Office of Emergency and Remedial Response, U.S Environmental Protection Agency, Washington, DC. December 21, 1993 [online]. Available http://www.epa.gov/oswer/riskassessment/pdf/irismemo.pdf [accessed May 13, 2008]. EPA (U.S. Environmental Protection Agency). 1993b. Provisional Guidance for Quanti- tative Risk Assessment of Polycyclic Aromatic Hydrocarbons. EPA/600/R-93/089. Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC. July 1993 [online]. Available: http://www.epa.gov/oswer/risk assessment/pdf/1993_epa_600_r-93_c89.pdf [accessed June 10, 2008]. EPA (U.S. Environmental Protection Agency). 1996a. Aroclor 1016 (CASRN 12674-11- 2). Integrated Risk Information System, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ncea/iris/subst/0462.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 1996b. Aroclor 1254 (CASRN 11097-69- 1). Integrated Risk Information System, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ncea/iris/subst/0389.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 1996c. PCBs: Cancer Dose-Response Assessment and Application to Environmental Mixtures. EPA/600/P-96/001F. Na- tional Center for Environmental Assessment, Office of Research and Develop- ment, U.S. Environmental Protection Agency, Washington, DC. September 1996 [online]. Available: http://www.epa.gov/pcb/pubs/pcb.pdf [accessed June 10, 2008]. EPA (U.S. Environmental Protection Agency). 1997a. Cumulative Risk Assessment Guidance-Phase I Planning and Scoping. Memorandum to Assistant Administra- tors, General Counsel, Inspector General, Associate Administrators, Regional Ad- ministrators, Staff Office Directors, from Carol M. Browner, Administrator, and Fred Hansen, Deputy Administrator, U.S. Environmental Protection Agency. July 3, 1997 [online]. Available: http://www.epa.gov/OSA/spc/pdfs/cumulrisk.pdf [ac- cessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 1997b. Guidance on Cumulative Risk Assessment. Part 1. Planning and Scoping. Science Policy Council, U.S. Environ- mental Protection Agency, Washington, DC. July 3, 1997 [online]. Available: http://www.epa.gov/OSA/spc/pdfs/cumrisk2.pdf [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 1997c. Polychlorinated Biphenyls (PCBs) (CASRN 1336-36-3). Integrated Risk Information System, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ncea/iris/subst/ 0294.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2000. Supplementary Guidance for Con- ducting Health Risk Assessment of Chemical Mixtures. EPA/630/R-00/002. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. Au- gust 2000 [online]. Available: http://www.epa.gov/NCEA/raf/pdfs/chem_mix/ chem_mix_08_2001.pdf [accessed June 10, 2008]. EPA (U.S. Environmental Protection Agency). 2001. Risk Assessment Guidance for Superfund, Volume 3- Part A: Process for Conducting Probabilistic Risk Assess- ment. EPA 540-R-02-002. Office of Emergency and Remedial Response, U.S. En- vironmental Protection Agency, Washington, DC. December 2001 [online]. Avail- able: http://www.epa.gov/oswer/riskassessment/rags3adt/ [accessed July 22, 2008].
Current Practice in Risk Assessment and Cumulative Risk Assessment 103 EPA (U.S. Environmental Protection Agency). 2002. Guidance on Cumulative Risk As- sessment of Pesticide Chemicals That Have a Common Mechanism of Toxicity. Office of Pesticide Programs, U.S. Environmental Protection Agency, Washing- ton, DC. January 14, 2002 [online]. Available: http://www.epa.gov/oppfead1/trac/ science/cumulative_guidance.pdf [accessed June 10, 2008]. EPA (U.S. Environmental protection Agency). 2003a. Human Health Toxicity Values in Superfund Risk Assessments. OSWER Directive 9285.7-53. Memorandum to Superfund National Policy Managers, Regions 1 - 10, from Michael B. Cook, Di- rector, Office of Superfund Remediation and Technology Innovation, Washington, DC. December 5, 2003 [online]. Available: http://www.epa.gov/oswer/riskassess ment/pdf/hhmemo.pdf [accessed May 13, 2008]. EPA (U.S. Environmental Protection Agency). 2003b. Framework for Cumulative Risk Assessment. EPA/630/P-02/001F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. May 2003 [online]. Available: http://oaspub. epa.gov/eims/eimscomm.getfile?p_download_id=36941 [accessed June 10, 2008]. EPA (U.S. Environmental Protection Agency). 2003c. The Feasibility of Performing Cumulative Risk Assessments for Mixtures of Disinfection By-Products in Drink- ing Water. EPA/600/R-03/051. National Center for Environmental Assessment, U.S. Environmental Protection Agency, Cincinnati, OH. June 2003 [online]. Available: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=56834 [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2005a. Guidelines for Carcinogen Risk Assessment. EPA/630/P-03/001F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. March 2005 [online]. Available: http://cfpub. epa.gov/ncea/cfm/recordisplay.cfm?deid=116283 [accessed Sept. 2, 2008]. EPA (U.S. Environmental Protection Agency). 2005b. Supplemental Guidance for As- sessing Susceptibility from Early-Life Exposure to Carcinogens. EPA/630/R- 03/03F. Risk Assessment Forum, U.S. Environmental Protection Agency, Wash- ington, DC. March 2005 [online]. Available: http://www.epa.gov/iris/children 032505.pdf [accessed June 16, 2008). EPA (U.S. Environmental Protection Agency). 2006. 1996 National Air Toxics Assess- ment Exposure and Risk Data. Technology Transfer Network National Air Toxics Assessment, U.S. Environmental Protection Agency [online]. Available: http:// www.epa.gov/ttn/atw/nata/ted/exporisk.html [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2007a. 1,1,1-Trichloroethane (CASRN 71-55-6). Integrated Risk Information System, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ncea/iris/subst/0197.htm [ac- cessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 2007b. Provisional Peer Reviewed Toxic- ity Values for Dimethyl Phthalate (CASRN 131-11-3). Superfund Health Risk Technical Support Center, National Center for Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, OH. September 25, 2007. EPA (U.S. Environmental Protection Agency). 2007c. Risk Management Decisions for Individual N-methyl Carbamate Pesticides. Pesticides, U.S. Environmental Protec- tion Agency [online]. Available: http://www.epa.gov/pesticides/cumulative/ carbamate_risk_mgmt.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2007d. 1996 National-Scale Air Toxics Assessment. Overview: EPAâs Use of Results. Technology Transfer Network, U.S.
104 Phthalates and Cumulative Risk Assessment: The Tasks Ahead Environmental Protection Agency [online]. Available: http://www.epa.gov/ttn/ atw/nata/ur.html [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2007e. 1996 National-Scale Air Toxics Assessment. Background on Risk Characterization. Technology Transfer Network, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ ttn/atw/nata/riskbg.html [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2007f. 1996 National-Scale Air Toxics Assessment. Summary of Results. Technology Transfer Network, U.S. Environ- mental Protection Agency [online]. Available: http://www.epa.gov/ttn/atw/ nata/risksum.html [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2007g. Concepts, Methods, and Data Sources for Cumulative Health Risk Assessment of Multiple Chemicals, Expo- sures and Effects: A Resource Document. EPA/600/R-06/013F. National Center for Environmental Assessment, Office of Research and Development, U.S. Envi- ronmental Protection Agency, Cincinnati, OH, in collaboration with U.S. Depart- ment of Energy, Argonne National Laboratory, Environmental Assessment Divi- sion, Argonne, IL. August 2007 [online]. Available: http://cfpub.epa.gov/ncea/ cfm/recordisplay.cfm?deid=190187 [accessed Nov. 12, 2008]. EPA (U.S. Environmental Protection Agency). 2008a. Target Compounds and Analytes. Superfund Analytical Services/Contact Laboratory Program, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/superfund/programs/ clp/target.htm [accessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 2008b. IRIS Glossary. Integrated Risk Information System (IRIS), U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/ncea/iris/help_gloss.htm#r [accessed July 23, 2008]. EPA (U.S. Environmental Protection Agency). 2008c. EPAâs Integrated Risk Information System: Assessment Development Procedures. IRIS Process (2008 Update). Na- tional Center for Environmental Assessment, U.S. Environmental Protection Agency. April 2008 [online]. Available: http://cfpub.epa.gov/ncea/cfm/record isplay.cfm?deid=190045 [accessed June 11, 2008]. EPA (U.S. Environmental Protection Agency). 2008d. Assessing Pesticide Cumulative Risk. Pesticides, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/pesticides/cumulative/index.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2008e. Common Mechanism Groups; Cumulative Exposure and Risk Assessment. Pesticides, U.S. Environmental Pro- tection Agency [online]. Available: http://www.epa.gov/pesticides/cumulative/ common_mech_groups.htm [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2008f. Proposed Approach for Estimation of Bin-Specific Cancer Potency Factors for Inhalation Exposure to Asbestos. Of- fice of Solid Waste and Emergency Response, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/oswer/riskassessment/asbestos/ pdfs/2008_prop_asbestos_approach.pdf [accessed July 22, 2008]. EPA (U.S. Environmental Protection Agency). 2008g. Revised N-Methyl Carbamate Cumulative Risk Assessment. Office of Pesticide Programs, U.S. Environmental Protection Agency [online]. Available: http://www.epa.gov/pesticides/cumulative/ common_mech_groups.htm#carbamate [accessed July 22, 2008]. Gustafson, J.B., J.G. Tell, and D. Orem. 1997. Selection of Representative TPH fractions Based on Fate and Transport Considerations. Total Petroleum Hydrocarbon Crite- ria Working Group Series, Vol. 3. Amherst, MA: Amherst Scientific Publishers
Current Practice in Risk Assessment and Cumulative Risk Assessment 105 [online]. Available: http://www.aehs.com/publications/catalog/contents/Volume3. pdf [accessed July 23, 2008]. ILSI (International Life Science Institute). 1999. A Framework for Cumulative Risk As- sessment, B. Mileson, E. Faustman, S. Olin, P. B. Ryan, S. Ferenc, and T. Burke, eds. Washington, DC: International Life Science Institute [online]. Available: http://rsi.ilsi.org/NR/rdonlyres/613A82AA-E74F-40B2-BA38- 1548A9C6AD8C/0/rsiframrpt.pdf [accessed July 23, 2008]. MADEP (Massachusetts Department of Environmental Protection). 2002. Characteriza- tion of Risks Posed by Petroleum Contaminated Sites: Implementation of MADEP VPH/EPH Approach. Final October 31, 2002. Policy No. WSC-02-411. Com- monwealth of Massachusetts Executive Office of Environmental Affairs, Depart- ment of Environmental Protection [online]. Available: http://www.mass.gov/ dep/cleanup/laws/02-411.pdf [accessed July 22, 2008]. MADEP (Massachusetts Department of Environmental Protection). 2003. Updated Petro- leum Hydrocarbon Fraction Toxicity Values for the VPH/EPH/APH Methodology. Office of Research and Standards, Massachusetts Department of Environmental Protection, Boston, MA. November 2003 [online]. Available: http://www.mass. gov/dep/cleanup/laws/tphtox03.pdf [accessed July 22, 2008]. Norberg, L., and G. WahlstrÃ¶m. 1988. Anaesthetic effects of flurazepam alone and in combination with thiopental or hexobarbital evaluated with an EEG-threshold method in male rats. Arch. int. Pharmacodyn. 292:45-57. OEHHA (Office of Environmental Health Hazard Assessment). 2008. Toxicity Criteria Database. Office of Environmental Health Hazard Assessment, California Envi- ronmental Protection Agency [online]. Available: http://www.oehha.ca.gov/ risk/chemicalDB//index.asp [accessed July 23, 2008]. Teuschler, L.K., G.E. Rice, C.R. Wilkes, J.C. Lipscomb, and F.W. Power. 2004. A feasi- bility study of cumulative risk assessment methods for drinking water disinfection by-product mixtures. J. Toxicol. Environ. Health Part A 67(8-10):755-777. Van den Berg, M., L. Birnbaum, A.T. Bosveld, B. BrunstrÃ¶m, P. Cook, M. Feeley, J.P. Giesy, A. Hanberg, R. Hasegawa, S.W. Kennedy, T. Kubiak, J.C. Larsen, F.X. van Leeuwen, A.K. Liem, C. Nolt, R.E. Peterson, L. Poellinger, S. Safe, D. Schrenk, D. Tillitt, M. Tysklind, M. Younes, F. Waern, and T. Zacharewski. 1998. Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Environ. Health Perspect. 106(12):775-792. Van den Berg, M., L.S. Birnbaum, M. Denison, M. De Vito, W. Farland, M. Feeley, H. Fiedler, H. Hakansson, A. Hanberg, L. Haws, M. Rose, S. Safe, D. Schrenk, C. Tohyama, A. Tritscher, J. Tuomisto, M. Tysklind, N. Walker, and R.E. Peterson. 2006. The 2005 World Health Organization reevaluation of human and mammal- ian toxic equivalency factors for dioxins and dioxin-like compounds. Toxicol. Sci. 93(2):223-241.