Low-Dose Risk Extrapolation

Dose-response assessment today is generally performed in two steps (1) assessment of observed data to derive a dose descriptor as a point of departure and (2) extrapolation to lower dose levels for the mmor type under consideration. The extrapolation is based on extension of a biologically based model (see Section 6.2.1) if supported by substantial data. Otherwise, default approaches that are consistent with current understanding of mode of action of the agent can be applied, including approaches that assume linearity or nonlinearity of the dose-response relationship, or both. The default approach is to extend a straight line to the human exposure doses. 

The outcome of low-dose extrapolation is the resulting lifetime cancer risk associated with estimated exposure for a particular population. A wide range of models have been developed for low-dose extrapolation of animal data to calculate a tolerable intake for an acceptable risk, often set at one extra cancer per million exposed persons (see Section 6.2.4 for acceptable risk). 

The multi-hit models are most suitable for extrapolating the effect of genotoxic substances. It is implicit in these models that aU hits occur in one specific cell that only begins to divide and develop into a tumor when it has received the necessary number of hits. However, this is in poor agreement with experimental data, which show that prohferation of the cells that have had their first hit (the initiated cells) into pre-neoplastic lesions considerably increases the risk of a second hit in an initiated cell. While the one-hit model often oversimplifies the process, the multi-hit models impose an unreasonable tight restriction of the possibdity of more than one critical hit affecting the same cell. 

According to the WHO (WHO/IPCS 1994, 1999 WHO 1996, 2000), it should be noted that cmde expression of risk in terms of excess incidence or numbers of cancers per unit of the population at doses or concentrations much less than those on which the estimates are based may be inappropriate, owing to the uncertainties of the quantitative extrapolation over several orders of magnitude. Estimated risks are therefore considered to represent only the plausible upper bounds and vary depending upon the assumptions on which they are based. 

In view of the considerable uncertainties in the extrapolation of results over several orders of magnitude, specification of risks in terms of predicted incidence or numbers of excess cancers per unit of the population implies a degree of precision that is considered misleading by some. Larsen (2006), e.g., noted that the model most often used in low-dose extrapolation is a linear extrapolation from the observable range, and the apparent precision of the calculations does not reflect the uncertainty in the risk estimate the results are therefore open to misinterpretation because the numerical estimates may be regarded as quantification of the actual risk. 

---

Quantitative extrapolation by mathematical modeling of the dose-response curve to estimate the risk at likely human exposures, i.e., low-dose risk extrapolation... 

In animal experiments exposures can be carefully controlled, and dose-response curves can be formally estimated. Extrapolating such information to the human situation is often done for regulatory purposes. There are several models for estimating a lifetime cancer risk in humans based on extrapolation from animal data. These models, however, are premised on empirically unverified assumptions that limit their usefulness for quantitative purposes. While quantitative cancer risk assessment is widely used, it is by no means universally accepted. Using different models, one can arrive at estimates of potential cancer incidence in humans that vary by several orders of magnitude for a given level of exposure. Such variations make it rather difficult to place confidence intervals around benefits estimations for regulatory purposes. Furthermore, low dose risk estimation methods have not been developed for chronic health effects other than cancer. The... 

Linear, no-threshold model (extrapolated upper bound on low-dose risk)... 

However, it does mean that extrapolation of a dose response in a linear fashion to zero could be too simplistic for some chemicals at least. Thus, mechanisms occurring after high-dose exposure may not be relevant to low-dose risk assessment. 

Mathematical models are used to extrapolate from animal bioassay or epidemiology data to predict low dose risk. Most models assume linearity with a zero threshold dose. 

In estimating the cumulative risk of a chemical in LCA, dose-response extrapolations can be based on toxicological benchmarks. Such a benchmark approach is considered more appropriate for use in comparative assessment contexts, such as in an LCA study. Benchmarks are an exposure measure associated with a consistent change in response, such as the 10% or even the 50% effect level. Regulatory-based measures do not necessarily provide a consistent risk basis for comparison, as they were often never developed for use in such a comparative context or to facilitate low dose-response extrapolation. Other data differences include the use of median, rather than extreme, data in the fate and exposure modeling, as well as the consideration of safety factors only as part of the uncertainty assessment and not as an integral part of the toxicological effects data. 

White, R. H., Cote, I., Zeise, L., Fox, M., Dominici, F, Burke, T. A., White, P. D., Hattis, D. B., and Samet, J. M. (2009). State-of-the-Science Workshop Report Issues and approaches in low-dose-response extrapolation for environmental health risk assessment. Environ Health Perspect 117(2), 283-287. 

Quantitative risk assessment requires extrapolation from results of experimental assays conducted at high dose levels to predicted effects at lower dose levels which correspond to human exposures. The meaning of this high to low dose extrapolation within an animal species will be discussed, along with its inherent limitations. A number of commonly used mathematical models of dose-response necessary for this extrapolation, will be discussed. Other limitations in their ability to provide precise quantitative low dose risk estimates will also be discussed. These include the existence of thresholds incorporation of background, or spontaneous responses modification of the dose-response by pharmacokinetic processes. 

Mantel-Bryan method was not proposed to provide necessarily valid estimates of low dose risk, but rather to provide "conservative" estimates of this risk. However, the "conservative" nature of this extrapolation methodology has been questioned by many authors (8-10). 

The US EPA has subsequently published a comprehensive toxicological review of bromate (US EPA, 2001). Studies with rats based on low-dose linear extrapolation, using the time-to-tumour analysis, and using the Monte Carlo analysis to sum the cancer potency estimates for kidney renal tubule tumoms, mesotheliomas, and thyroid follicular cell tumours, gave an upper-bound cancer potency estimate for bromate ion of 0.70 per mg/kg day. This potency estimate corresponds to a drinking water unit risk of 2 x 10 per pg/L, assuming a daily water consumption of 2 litres/day for a 70-kg adult. Lifetime cancer risks of 10 , 10 , and 10 are associated with bromate concentrations of 5, 0.5, and 0.05 pg/L, respectively. A major source of uncertainty in these estimates is from the interspecies extrapolation of risk from rats to humans. 

One important point of controversy in risk extrapolation is the existence of the threshold level for carcinogenic and mutagenic response to a pollutant. Some argue that an organism is able to cope with low doses of a substance through metabolic processes or repair mechanisms, so that harmful effects do not appear until a certain minimum threshold, or "safe dose", is surpassed. Others contend that a carcinogenic substance must be considered potentially harmful at any dose, and that even a single molecule may initiate a tumor at the cellular level. This is the so-called "one-hit" hypothesis. 

There have been a number of recent survey articles and theoretical papers describing the available models for low-dose extrapolation. Through a literature review the most prominent models have been selected and discussed below. However, there are other models, less commonly used, that were not mentioned here for the sake of brevity. The models addressed below represent a good cross-section of the different features and capabilities that are pertinent to carcinogenic risk estimation. 

This model tends to approach a zero probability rapidly at low doses (although it never reaches zero) and thus is compatible with the threshold hypothesis. Mantel and Bryan, in applying the model, recommend setting the slope parameter b equal to 1, since this appears to yield conservative results for most substances. Nevertheless, the slope of the fitted curve is extremely steep compared to other extrapolation methods, and it will generally yield lower risk estimates than any of the polynomial models as the dose approaches zero. 

In radiation protection and risk assessment linear extrapolations are commonly used. Linearity, however, is not based on experimental evidence in the very low dose range and a departure from it would imply an over- or underestimation of risks. There are both theoretical and experimental indications wich contradict the linearity at doses considered here (e.g. Brown, 1977 ... 

Resmethrin (I) Likely to be carcinogenic to humans based on increased incidences of benign and malignant liver tumors in female rats and male mice. A low-dose extrapolation approach was applied to the experimental animal data in order to estimate human cancer risk [100]. No oncogenic effects were seen [101]. 

---