Health Risks from Dioxin and Related Compounds Evaluation of the EPA Reassessment (2006) / Chapter Skim
Currently Skimming:
2 General Considerations of Uncertainty and Variability, Selection of Dose Metric, and Dose-Response Modeling
2 General Considerations of Uncertainty and Variability, Selection of Dose Metric, and Dose-Response Modeling
Pages 45-74
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 45... ...
49) Significant uncertainties remain in understanding human health risks from 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
|
From page 46... ...
This chapter identifies the major categories of decisions that analysts generally make when developing risk estimates in the context of the four traditional steps of risk assessment: hazard identification and classification, exposure assessment, dose-response assessment, and risk characterization (NRC 1983)
|
From page 47... ...
This generic approach results in limited discussions of variability and uncertainty. The committee found that the lack of a specific context and absence of a focused exposure assessment that would link sources to potential health effects in individuals, or in the population, severely limited both EPA's and the committee's abilities to appropriately characterize variability and uncertainty in risk estimates related to exposure to TCDD, other dioxins, and DLCs.
|
From page 48... ...
EXPOSURE ASSESSMENT EPA provided the committee with an updated exposure inventory (EPA 2005b) , which provides an extensive review of the existing database of exposure data for TCDD, other dioxins, and DLCs.
|
From page 49... ...
Yet another area of uncertainty is determination of background levels when many samples lie below the analytical limit of detection. This issue arises in any exposure assessment, and several widely used options address it (e.g., assume all nondetects are true zeroes, assign a value of either /1 2 or 1 times the detection limit, or fit a distribution to the data)
|
From page 50... ...
. The exposure assessment also demonstrates that TCDD intake for children based on age-specific food consumption and average food concentrations exceeds adult intake estimates on a bodyweight basis (although their intake on a mass basis is lower)
|
From page 51... ...
The Reassessment suggests that this discrepancy arises from the presence of an historical body burden and lipid concentration, but it does not consider other possibilities. GENERAL ISSUES RELATED TO VARIABILITY AND UNCERTAINTY ASSOCIATED WITH SELECTION OF DOSE METRIC AND DOSE-RESPONSE MODELING EPA makes a number of assumptions about the appropriate dose metric and mathematical functions to use in the Reassessment's dose-response analysis (see "Selection of Dose Metric" and "Dose-Response Modeling" in this chapter for specific issues related to dose metric and dose-response modeling)
|
From page 52... ...
52 01 of 01 05 10 ED ED ED LED Other Point Departure · · · · · the daily average under burden Metric Risk Average dose Area curve Lifetime body Peak Other Dose · · · · · Cancer Low-dose linear Nonlinear Multiple Other Dose-Response Model · · · · Characterizing in ED. Set studies on from Faced Data limit multiple EPA Choose individual Use studies Bioassay · · confidence Decisions lower from Key multiple LED, Set of Choose individual studies Use studies dose; Epidemiological Data · · effective data Categories ED, data 2-1 bioassay for Epidemiological and Epidemiological data Bioassay Other TABLE Basis Quantification · · · · Abbreviations:
|
From page 53... ...
53 and dose; Choice factor factor. change nervous ainty Effect effective (Intraspecies)
|
From page 54... ...
approach to derive a POD for noncancer end points. Although a lower confidence bound on an ED was cited in the literature to define a BMD, EPA's BMD guidance document (EPA 2000b)
|
From page 55... ...
The risk estimates can be most fully characterized by performing probabilistic analyses when possible and by presenting the range of possible risk estimates rather than by reporting the single point estimates. Risk characterization should provide useful information to risk managers to help them understand the variability and uncertainty in the risk estimates.
|
From page 56... ...
For dioxin, these issues are best illustrated in relation to the estimation of cancer risk. The choice of one possible approach, linear extrapolation from a POD, results in a CSF that could be used to estimate the lifetime cancer risk for the U.S.
|
From page 57... ...
, plasma or tissue concentrations, body burden, and function-related biomarkers of the internal dose such as aromatic hydrocarbon receptor (AHR) occupancy or changes in cytochromes P450A1/2 protein (CYP1A1/ 2)
|
From page 58... ...
. Body burdens after shorter periods of administration (non-steady state)
|
From page 59... ...
This assumption represents a reasonable default because the body burden generally appears to be proportional to tissue concentration, with some caveats noted in Chapter 5, and the toxic effects of TCDD, other dioxins, and DLCs increase with increased tissue concentration. However, the use of body burden as a dose metric (or a dose metric based on tissue concentration)
|
From page 60... ...
-- for example, due to differences in body fat content. Because half-life depends on both CL and V, and body fat content represents the major determinant of V for TCDD and other dioxins, a species with a proportionately higher body fat content would have a proportionately higher value of V, a proportionately longer half-life, and greater body burden at steady state for the same daily intake.
|
From page 61... ...
Differentiation of free compound from lipid-bound compound within a PBPK model could provide the most relevant dose metric for dose-response assessment. The approximately 100-fold difference between rats and humans in TCDD half-life combined with Equation 2-1 suggests that a 100-fold lower daily intake in humans yields a total body burden equal to that in rats (assuming the same bioavailability)
|
From page 62... ...
The Reassessment applies the same body burden correction factor between rats and humans for liver cancer and for nonhepatic effects. The proportionately higher concentrations in the livers of rats compared with humans means that a proportionately higher daily intake would be necessary in humans to produce a comparable hepatic concentration.
|
From page 63... ...
The same PBPK model might also be used to explore the influence of human variability in body composition on the elimination halflife and therefore the body burden at steady state. The Reassessment did not consider this approach or quantify its impact, despite its recognition of tissue concentration as the best dose metric.
|
From page 64... ...
. In the case of TCDD, other dioxins, and DLCs, it is important to assess the population-based dose-response relationship for cancer, birth defects, immunotoxic effects, and so forth.
|
From page 65... ...
Here, R(d) describes the mean response level of the toxicological outcome (e.g., cognitive function as measured in terms of IQ test score in the case of exposure to a neurotoxin)
|
From page 66... ...
. That approach first identifies an adverse response level, which demarcates normal and abnormal (or adverse)
|
From page 67... ...
In the present context, statistical power refers to the general ability of an experiment, and its associated data set, to provide information needed to make a reliable inference, including testing positive dose effects and ascertaining a fitted dose-response model. The Reassessment did not discuss the issue of statistical power, although the cancer guidelines (EPA 2005a, see also Appendix B)
|
From page 68... ...
to account for both nonlinear and linear shapes of the dose response for noncancer effects. However, the committee recommends that EPA apply similar efforts in dose-response modeling of human cancer data (see Chapter 5)
|
From page 69... ...
estimates from `good' and `marginal' model fits were subjectively evaluated for stability and biological plausibility with regard to the observed data. This evaluation identified some potential problems with some of the Vmax estimates.
|
From page 70... ...
With limited data (e.g., about three dose groups for noncancer data) and limited statistical power, many of the data sets (including epidemiological studies)
|
From page 71... ...
, an approach recommended in the BMD guidance document (EPA 2000b) and implemented in EPA's BMD software program.
|
From page 72... ...
For example, many of the data sets of noncancer effects yielded a Hill coefficient greater than 1.5, indicating a plausible nonlinear dose response. However, those studies lacked adequate statistical power to estimate the Hill coefficient reliably, rendering the estimate statistically nonsignificant (that is, the confidence interval includes unity)
|
From page 73... ...
. The committee recommends that EPA use simple PBPK models to define the magnitude of any differences between humans and rodents in the relationship between total body burden at steady-state concentrations (as calculated from the intake, half-life, bioavailability)
|
From page 74... ...
74 HEALTH RISKS FROM DIOXIN AND RELATED COMPOUNDS · The committee notes that EPA would substantially improve its transparency and management of the complexity of the risk assessment of TCDD, other dioxins, and DLCs by creating an ongoing process for clearly identifying and updating the key assumptions that support the quantitative risk assessment. This process would essentially require viewing the risk assessment as an ongoing and iterative effort in which EPA continues to create incentives to obtain and use better information when possible and appropriate.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.