Dose response extrapolation from

The situation becomes bleaker still when we recognize that, for a given substance, there may be many different conditions A under which its adverse effects and their dose-response characteristics have been investigated Results may be available from several different epidemiological studies, in different groups exposed under different circumstances, and with results that are not entirely consistent with each other. Some of the conditions may involve experimental data, similarly variable in outcome and in how they will be interpreted by different scientists. So, when we are faced with toxic hazard and dose-response data from studies involving conditions A1 through A12, which, if any, are most useful and relevant for extrapolation to condition B ... 

In some cases, the duration of exposure that is or might be experienced by the population of interest might not match that involved in the study. So, for example, dose-response information from relatively short-term exposures might in some cases be the only available information when the concern is long-term, or even lifetime exposure in the population that is the subject of the risk assessment. If the risk assessment is to be completed before new, long-term data can be developed, some justification will have to be found for extrapolation of the short-term data to estimate the consequences of long-term exposure. 

If dose-response data from an animal study are selected as the principal basis for assessing risk, how are they to be applied (extrapolated) to the human population that is the subject of the risk assessment This is the problem of interspecies extrapolation. 

Statistical models. A number of statistical dose-response extrapolation models have been discussed in the literature (Krewski et al., 1989 Moolgavkar et al., 1999). Most of these models are based on the notion that each individual has his or her own tolerance (absorbed dose that produces no response in an individual), while any dose that exceeds the tolerance will result in a positive response. These tolerances are presumed to vary among individuals in the population, and the assumed absence of a threshold in the dose-response relationship is represented by allowing the minimum tolerance to be zero. Specification of a functional form of the distribution of tolerances in a population determines the shape of the dose-response relationship and, thus, defines a particular statistical model. Several mathematical models have been developed to estimate low-dose responses from data observed at high doses (e.g., Weibull, multi-stage, one-hit). The accuracy of the response estimated by extrapolation at the dose of interest is a function of how accurately the mathematical model describes the true, but unmeasurable, relationship between dose and response at low doses. 

Once an assessment has determined that the data indicate human risk potential for reproductive and developmental toxicity, the next step is to perform a quantitative evaluation. Dose-response data from human and experimental animal reproductive and developmental toxicity studies are reviewed to identify a no-observed-adverse-effect level (NOAEL) or a lowest-observed-adverse-effect level (LOAEL), and/or to derive a benchmark dose (BMD). Duration adjustments of the NOAEL, LOAEL, or BMD are often made, particularly for inhalation exposures when extrapolating to different exposure scenarios. Such adjustments have not been routinely applied to developmental toxicity data. The subcommittee recommends that duration adjustments be considered for both reproductive and developmental toxicity... 

Although there are dose-response data from an animal inhalation exposure study (McNamara et al., 1975, see Section 5.1.1), route-to-route extrapolation (from inhalation to oral, as calculated in Section 5.2.1) is not considered appropriate because the exposure protocol of McNamara et al. (1975) resulted in rat skin tumors which might have occurred, not a result of systemic uptake, but as a result of dermal contact with sulfur mustard vapor (perhaps trapped by the rat pelt). Therefore, there is no method for estimating the dermal dose of sulfur mustard, or for converting this to an oral dose. 

Figure 5. Hypothetical dose-response curves for chemically-induced carcinogenicity, showing measured dose-response curves from three studies (top right-hand quadrant) and some possible ways those curves might behave in the low dose-low risk region (lower left-hand quadrant, in The Range of Extrapolation). Note The graph is not drawn to scale. The lower left-hand quadrant has been greatly expanded to show the possibilities for extrapolation. See text, pages 174 to 179 for a full discussion of this graph.

Table IV. Cougarison of Virtually Safe Doses (VSD) Leading to an Excess Risk of 10 for Various Dose-Response Extrapolation Models (models applied to data from (41)...

There remains a paucity of data on other nerve agents and more complete dose-response relationships from animal studies which optimize extrapolation of animal-derived data to man are needed. 

Polycyclic aromatic hydrocarbons are capable of forming adducts with DNA in cells. Exposure to PAHs from creosote were measured in the personal work areas of coke oven workers in the Czech Republic (Lewtas et al. 1997). Measured levels of DNA adducts in white blood cells of a nonoccupationally exposed population were well correlated with the low to moderate environmental exposures. The DNA adducts of the coke oven workers who were exposed to carcinogenic PAHs at levels of <5->200,000 ng/m3 (<0.005->200 pg/m3) did not correlate well with the exposure levels. These authors concluded that various mechanisms were responsible for the lower DNA-binding potency at the higher exposure levels, precluding the use of a linear model for dose-response extrapolation in risk assessment. 

NOAEL (no-observed-adverse-effect level) is defined as the highest dose at which no adverse effects are observed in the most susceptible animal species. The NOAEL is used as a basis for setting human safety standards for acceptable daily intakes (ADIs), taking into account uncertainty factors for extrapolation from animals to humans and inter-individual variabilities of humans. The adequacy of any margin of safety or margin of exposure must consider the nature and quality of the available hazard identification and dose-response data and the reliability and relevance of the exposure estimations. In some cases, no adverse endpoint can be identified such as for many naturally occurring compounds that are widespread in foods. In that case, an ADI Not Specified is assigned. ... 

Zero-Threshold Linear Hypothesis—The assumption that a dose-response curve derived from data in the high dose and high dose-rate ranges may be extrapolated through the low dose and low dose range to zero, implying that, theoretically, any amount of radiation will cause some damage. 

The following example is based on a risk assessment of di(2-ethylhexyl) phthalate (DEHP) performed by Arthur D. Little. The experimental dose-response data upon which the extrapolation is based are presented in Table II. DEHP was shown to produce a statistically significant increase in hepatocellular carcinoma when added to the diet of laboratory mice (14). Equivalent human doses were calculated using the methods described earlier, and the response was then extrapolated downward using each of the three models selected. The results of this extrapolation are shown in Table III for a range of human exposure levels from ten micrograms to one hundred milligrams per day. The risk is expressed as the number of excess lifetime cancers expected per million exposed population. 

Steinhausler (1987) and Martell (1987) review the dosimetric models and related model studies. Their view is that there are still very large uncertainties in the existing data and in the extrapolation from the exposure and response data for underground miners and experimental animals to the health effects of the radon progeny levels to which the general public is exposed. B.L. Cohen (1987) describes his work to relate radon measurements with lung cancer rates for various geographical areas to test the concept of a dose threshold. 

---