Predictive exposure models

In this chapter the new trends in analytical chemistry for determining classical and emerging pollutants, as well as the use of predictive exposure models have been reviewed and their respective benefits and shortcomings have been briefly discussed. 

Besides the LCA approach, also risk assessment can be performed analysing the chemical compounds or modelling via predictive exposure models. Both types of approaches have their justification to measure environmental concentrations of chemicals in the environment with laboratory measurement is still the most reliable way for determination. But it goes along with the disadvantage of high investments concerning time and money. Besides that laboratory approaches are limited in terms of space and time, and in consequence, the survey of many micro-pollutants and their... 

Use of Predictive Exposure Models in Risk Assessment 200 Probabilistic Approach 201... 

IMP (1986). UK Predictive Exposure Model (POEM) Estimation of Exposure and Absorption of Pesticides by Spray Operators, UK Scientific Sub-committee on Pesticides and British Agrochemical Association Joint Medical Panel, Pesticides Safety Directorate, York, UK. 

There is also the question of range of applications for EPA s lEUBK and similar models. Few predictive exposure models for any substance are expected to be equally applicable for all environmental contamination settings for that matter, neither is any model likely to be universally useful or adaptable. One obvious reason for any model s limits is the restricted scope of the empirical data typically used to evaluate models. The usual process is to evaluate models with data most directly relevant to the intended applica-tion(s). One then has the narrower task of determining whether a model such as the lEUBK model is appropriate to its actual uses. 

The SWRRB runoff model coupled to the EXAMS fate model can be used to predict exposure levels of chemicals to aquatic organisms. Safety factors can then be calculated. 

The next step is impact prediction that requires detailed quantitative information about the sources of risk agents, exposure models, the receptors and possible changes in the state of these receptors caused by the defined agents. If the CLL concept was selected for assessment ecosystem effects it should firstly be utilized for impact baseline studies or assessing the do-nothing scenario. In this context CLL calculation includes the following steps (Bashkin, 2002) ... 

The prediction of a given contour shape, i.e., resist profile, requires both exposure and development models. We will first examine the various exposure models which have been developed and then combine these with development studies in order to predict resist profiles and compare them with experiment. 

For risk assessment of chemicals based on a DNEL as described above, the output of the exposure assessment (Chapter 7) is usually an estimate of dose or concentration. Exposure data can either be measured or predicted. Measured exposure data are preferred if they are valid, i.e., an actual exposure estimate. However, in most cases, measured data are not available and therefore model-generated data must be used for the risk characterization, i.e., a predicted exposure estimate. 

Predictive methods of exposure assessment often rely on single values for input parameters to the exposure model that represent one point on the distribution curve of all possible values for this parameter. This point value can range from a 50th percentile, mean, median, or typical value to a worst-case estimate. In the predictive exposure assessment, a number of parameters are integrated through an algorithm to produce an output such as the predicted environmental concentration (PEC). If many worst-case values are involved, this integration can result in a PEC that has a... 

Finally, the MOS should also take into account the uncertainties in the estimated exposure. For predicted exposure estimates, this requires an uncertainty analysis (Section 8.2.3) involving the determination of the uncertainty in the model output value, based on the collective uncertainty of the model input parameters. General sources of variability and uncertainty in exposure assessments are measurement errors, sampling errors, variability in natural systems and human behavior, limitations in model description, limitations in generic or indirect data, and professional judgment. 

An example of an exposure model to predict distribution in environmental media and estimation of the proportion of total exposure by various routes from consumer products is the BUSES (Section 7.2.4.3). It is important to recognize that the proportions of total intake from various media may vary, based on circumstances. 

Absorption, Distribution, Metabolism, and Excretion. Levels of cresols in blood were obtained from a single case report of a dermally exposed human (Green 1975). Data on the toxicokinetics of cresols in animals were contained in two acute oral studies that provided only limited quantitative information on the absorption, metabolism, and excretion of cresols (Bray et al. 1950 Williams 1938). A more complete oral toxicokinetics study, in addition to studies using dermal and inhalation exposure, would provide data that could be used to develop predictive pharmacokinetic models for cresols. Inclusion of several dose levels and exposure durations in these studies would provide a more complete picture of the toxicokinetics of cresols and allow a more accurate route by route comparison, because it would allow detection of saturation effects. Studies of the tissue distribution of cresols in the body might help identify possible target organs. 

First-order error analysis is a method for propagating uncertainty in the random parameters of a model into the model predictions using a fixed-form equation. This method is not a simulation like Monte Carlo but uses statistical theory to develop an equation that can easily be solved on a calculator. The method works well for linear models, but the accuracy of the method decreases as the model becomes more nonlinear. As a general rule, linear models that can be written down on a piece of paper work well with Ist-order error analysis. Complicated models that consist of a large number of pieced equations (like large exposure models) cannot be evaluated using Ist-order analysis. To use the technique, each partial differential equation of each random parameter with respect to the model must be solvable. 

For a better idea of the toxicity of VOCs, we can look more closely at some studies of TCE (Bogen et al., 1998). In vitro uptake of C-14-labeled trichloroethylene (TCE) from dilute (similar to 5-ppb) aqueous solutions into human surgical skirt was measured using accelerator mass spectrometry (AMS). The AMS data obtained positively correlate with (p approximate to 0) and vary significantly nonlinearly with (p = 0.0094) exposure duration. These data are inconsistent (p approximate to 0) with predictions made for TCE by a proposed EPA dermal exposure model, even when uncertainties in its recommended parameter values for TCE are considered but are consistent (p = 0.17) with a 1-compartment model for exposed skin-surface. This study illustrates the power of AMS to facilitate analyses of contaminant biodistribution and uptake kinetics at very low environmental concentrations. Eurther studies could correlate this with toxicity. 

Internal burdens of epoxybutene in humans resulting from exposure to butadiene were predicted from models by Kohn and Melnick (1993), Johanson and Filser (1996) and Csanady et al. (1996) and were compared with simulations for rats and mice. In the model of Kohn and Melnick (1993), metabolic parameters were incorporated which had been obtained by Csanady et al. (1992) by measuring butadiene and epoxybutene oxidation and epoxybutene hydrolysis in human liver and lung microsomes in vitro, and conjugation of epoxybutene with glutathione in human liver and lung cytosol. Tissue blood partition coefficients were theoretically derived. The body burden of epoxy butene following exposure to 100 ppm butadiene for 6 h was predicted to be 0.056 pmol/kg in humans. 

LAS. LAS environmental levels in United States surface water are expected to range from <0.04 to 0.14 mg/L, as predicted from modeling results and from actual monitoring data (see Exposure Assessment section). Because of the agreement between results from the national LAS monitoring efforts and model predictions, there is a high level of confidence in this exposure assessment. 

Mathematical construction of physical/ chemical processes that predict the range and probability density distribution of an exposure model outcome (e.g. predicted distribution of personal exposures within a study population)... 

If field studies are conducted in which the variables are clearly delineated, it is possible that a model could be developed that would predict the maximum exposure under registered use conditions, and such a model would be of great value in the assessment of hazard to workers. This approach may be feasible for the orchard scenario because of the large number of studies that have been carried out. Some of the factors which should be taken into consideration in developing an exposure model and areas where there are insufficient data to make an accurate estimate will be discussed. 

Other published works relevant to estimation of properties and reactivity of chemicals may be of value to the reader. A book edited by Neely and Blau (1985) (Environmental Exposure from Chemicals) is less comprehensive but has several excellent chapters on basic information, such as the one on biodegradation by Klecka. Other volumes that may be useful include The Properties of Drugs, edited by Yalkowsky et al. (1980) and Aqueous Solubility by Yalkowsky and Banerjee (1991). Chemical Exposure Predictions, edited by Calamari (1993), is somewhat different in focus but may be useful for information on exposure modeling, especially for the soil compartment. 

Fig. 13.3 D ose proportionality assessment. The solid line is the predicted exposure based on a linear mixed effects power model having a reference subject. For the tasidotin plots, the predicted line is for a reference patient having a BSA of 1.83 m2.

---