Quantitative Risk Estimates

Although it seems clear that animal tests can be used with a reasonable degree of confidence to identify chemicals with carcinogenic activity, the application of such test data in the quantitative estimation of human risk is a more difficult process. The formulation of risk estimates requires extrapolations fi om effects observed in different species which are generally exposed to much higher dosage levels them those to which the human population would be ex- 

Comparative Dose Response Data There are only a limited number of cases in which it has been possible, even on a crude level, to compare animal and human dose responses to the same carcinogens. The paucity of data in this area is related to the fact that in most epidemiological studies the actual doses of the toxic substance is not known with certainty and that while complete dose response curves can be obtained in animal studies, such studies are very expensive. A systematic attempt to estimate human exposures and compare human and animal dose response relationships for several known carcinogens has been made and is contained in a report by the National Academy of Sciences and National Research Council on Environmental Studies Board on Pest Control.The carcinogens for which responses were compared in the study were benzidine chlornaphazine diethylstilbestrol aflatoxin Bl vinyl chloride, and cigarette smoke. 

On the basis of this albeit limited data, the Meselson Committee, which conducted the above study, has suggested as a working hypothesis that .. . . in the absence of countervailing evidence for the specific agent in question, it appears reasonable to assume that the lifetime cancer incidence induced by chronic exposure in man can be approximated by the lifetime incidence induced by similar exposure in laboratory animals at the same total dose per body weight . 

There have been several mathematical models developed for estimating the effects of exposures at dosage levels below those for which test data are available. It is beyond the scope of this chapter to review these models in any detail and for a more detailed discussion the reader is referred to reviews such as those by Schneiderman et al. Cranmer, and Hoel et al. What will be discussed here are the biological bases and implications of some of these models, and a recently conducted, large-scale experiment which attempted to determine the shape of the dose response curve for mice fed low levels of the carcinogen 2-acetylaminofluorene (2-AAF). 

There are basically two types of mathematical models commonly used to represent dose response relationships. One type which is referred to as a dichotomous or yes-no modelis concerned with whether or not a specific response, such as cancer, has occurred with increased incidence in the treated pwpulation. The second type, which is a time-to-response model, attempts to relate dose levels to the time of appearance of the measured effect. 

---

The model contains a surface energy method for parameterizing winds and turbulence near the ground. Its chemical database library has physical properties (seven types, three temperature dependent) for 190 chemical compounds obtained from the DIPPR" database. Physical property data for any of the over 900 chemicals in DIPPR can be incorporated into the model, as needed. The model computes hazard zones and related health consequences. An option is provided to account for the accident frequency and chemical release probability from transportation of hazardous material containers. When coupled with preprocessed historical meteorology and population den.sitie.s, it provides quantitative risk estimates. The model is not capable of simulating dense-gas behavior. 

Dispensing with an extrapolation. The data quality in these cases are so poor that consideration should be given to dispensing with a quantitative risk estimation altogether. 

The health risks associated with ozone depletion will principally be those due to increased ultraviolet-B (UV-B) radiation in the environment, that is, increased damage to the eyes, the immune system, and the skin. Some new risks may also be introduced with the increased use of alternatives to the ozone-depleting substances (ODS). Quantitative risk estimates are available for some of the UV-B-associated... 

The Monographs represent the first step in carcinogenic risk assessment, which involves examination of all relevant information in order to assess the strength of the available evidence that certain exposures could alter the incidence of cancer in humans. The second step is quantitative risk estimation. Detailed, quantitative evaluations of epidemiological data may be made in the Monographs, but without extrapolation beyond the range of the data available. Quantitative extrapolation from experimental data to the human situation is not undertaken. 

The media never made an effort to educate the public on quantitative risk estimates and use these to put nuclear risks into perspective. I published papers on catalogs of risks,10 and used them constantly in public presentations and interviews with reporters. These are easily understandable to anyone, butthey were never transmitted by the media. 

Because genetic effects in man cannot be reliably predicted from submammalian tests, smother alternative for quantitative risk estimation is to rely on data on the mouse. Despite reservations about the sensitivity of mouse tests and the statistical power of experiments with small numbers of animals, an in vivo mammalian system incorporates the more relevant biologic and metabolic factors necessary to make quantitative human risk estimates. 

UNSCEAR473 based its quantitative risk estimates mainly on severe dominant diseases. Recessive and complexly inherited diseases were regarded as too poorly understood and too diluted by time to be considered. The BEIR reports302 303 included disease of complex inheritance by making arbitrary assumptions about the extent to which the incidence is mutationally caused and about the time distribution. We agree with the policy of placing major emphasis on severe dominant mutations, but will treat the subject in a more systematic way. 

If carcinogenicity data are unavailable or inconclusive, the next step (Level 4) is to test for mutagenicity in the mouse. In many cases, the presence or absence of mutagenicity in the short-term tests and in the mouse may be sufficient evidence for a decision. In the most difficult cases, a quantitative risk estimate may be needed. 

This subsection discusses practical approaches to assessing uncertainty in Superfiind site risk assessments and describes ways to present key information bearing on tlie level of confidence in quantitative risk estimates for a site. The risk measmes used in Superfund site risk assessments usually are not fully probabilistic estimates of risk, but conditional estimates given a considerable number of assumptions about exposure and toxicity (e.g., risk given a particular future land use). Thus, it is important to fully specify the assumptions and uncertainties inlierent in tlie risk assessment to place tlie risk estimates in proper perspective. Another use of uncertainty characteriztition can be to identify areas where a moderate amount of additional data collection might significantly improve the basis for selection of a remedial alternative. 

Starr TB, Gibson JE. 1985. Tire mechanistic toxicology of fonnaldehyde and its implications for quantitative risk estimation. Annu Rev Pharmacol Toxicol 25 745-767. 

In conclusion, early risk evaluation is an absolutely essential feature or element in the development of new products. The range and sequence of tests to be done at this stage are subject to discussion and various approaches can be chosen. A first attempt to establish a structured risk evaluation scheme was made by Gisi and Staehle-Czech (.13). This is an interesting and valuable approach towards a more quantitative risk estimation. 

For example, the pesticide fosetyl al (Aliette) is a fungicide proposed for use on hops. Many fungicides have been shown to induce tumors in animals and are considered to present oncogenic risks to humans. Aliette caused an increase in bladder tumors in rats, but only at very high dose levels (40,000 ppm). Aliette was classified as a group C carcinogen, and a quantitative risk estimate demonstrated upper-bound worst-case risks were less than one in one hundred million. 

Stanfill, S.B. and D.L. Ashley Quantitation of flavor-related aUcenylbenzenes in tobacco smoke particulate by selected ion monitoring gas chromatography-mass spectrometry J Agr. Food Chem. 48 (2000) 1298-1306. Starr, T.B. and J.E. Gibson The mechanistic toxicology of formaldehyde and its implications for quantitative risk estimation Ann. Rev. Pharmacol. Toxicol. 25 (1985)745-767. 

Using the consequence and likelihood categories, risk matrix, and risk evaluation criteria, the team reviewed three release scenarios (small, medium, and large) for the segments identified for each of the chemical movements. The result of the semi-quantitative risk estimation for this facility s hazardous material transportation operation is detailed in Table 4.12. From this results table, the following are determined ... 

Quantitative risk estimate deveioped for baseline and options to reduce risk... 

Risk estimation is projecting frequency and severity, the same tasks others include in risk analysis. This diagram infers quantitative risk estimation using probabilities to estimate the likelihood of a hazardous event. It seeks to place values on the severity. 

As a result, quantitative risk estimates should be expressed in terms of probability distributions rather than being presented as point estimates. However, one of the technical... 

Risk description the semi-quantitative risk estimation is a crucial element to reach the goals of the method. In such an estimation method, verbal statements are used to describe the risk factors. According to these verbal statements, a value on a numerical scale is assigned for each... 

The next three subsections describe the role that quantitative risk estimates played in addressing and resolving three important regulatory issues Anticipated Transients Without Scram, Auxiliary Feedwater System Reliability, and Station Blackout. Following these discussions, current policies and practices of the NRC regarding the use of quantitative risk estimates are discussed in subsections addressing the Safety Goal Policy, the Backfit Rule, and Individual Plant Examinations. 

---