3. Dose-Response Evaluation
Dose-response evaluation is the process of quantitatively evaluating the toxicity information and characterizing the relationship between the dose of a chemical and the incidence of specific biological effects in the exposed population. Dose-response evaluation as applied here identifies the toxicity values which will be used in risk characterization, and briefly describes the sources of information and how the values were derived. Detailed information regarding derivation of the toxicity values is available in the sources cited, and a broad overview of dose-response for the POPCs may be found in the RAND literature review.[502] The toxicity values are used in the risk characterization step to estimate the likelihood of specific biological effects occurring in humans at the different exposure levels.
Appropriate toxicity values were compiled or extrapolated from published sources. Most of the published values are associated with a medium to high level of confidence; however, one may assume the extrapolated values to be associated with a low level of confidence.
The toxicity values used here are of two types: toxicity values used to evaluate noncarcinogenic effects, and toxicity values used to evaluate carcinogenic effects. In the risk characterization, reference doses (RfDs) are used to evaluate potential noncarcinogenic effects for various exposure durations, as follows:
The exposure durations as defined above are consistent with definitions the EPA has used in the past. Recent EPA risk assessments (e.g., for chlorpyrifos[503] ), however, employ a somewhat different convention (not used for this HRA), as follows: acute (1 day), short-term (1-30 days), intermediate term (30 days to several months), and chronic (several months to lifetime).
Slope factors are used to evaluate potential carcinogenic effects. In some cases where published toxicity values were not available, provisional values were derived by extrapolation as described later.
a. Noncarcinogenic Effects
For purposes of this HRA two sets of toxicity values are available for the evaluation of noncarcinogenic effects, referred to as follows: "standard toxicity values" (the toxicological approach), and "other human benchmarks" (the epidemiological approach). The standard toxicity values, compiled here, comprise those that have typically been used in conducting quantitative risk assessments for many years. Ideally, the standard values are values verified by the EPA, and based on controlled laboratory studies. They are usually based on animal studies, but may be based on human studies. The other human benchmarks, provided in Tab J, are not typically used in quantitative risk assessment, but serve a useful role in relation to the HRA. The other benchmarks are based on human studies identified in the literature, such as worker exposure studies.
(1) Standard Toxicity Values
Tables 92 through 100 present the standard toxicity values and associated information for the evaluation of noncarcinogenic effects. The tables list the following published values:
The RfD is the key toxicity value which is used in the risk characterization to judge potential noncarcinogenic effects. For each combination of chemical, route, and exposure duration, an RfD has been identified wherever possible and appropriate. In each circumstance, the first value listed (NOEL, NOAEL, or LOEL) is the value upon which the RfD is based. The EPA typically takes the most useful "effect level" identified (e.g., NOAEL) and applies a combination of uncertainty and modifying factors to derive the RfD. In most cases listed in Tables 92 through 100, the combined factor ranges between 100 and 300. Given that the RfDs are usually based on some form of no-effect level, they likely err on the side of caution.
An RfD is defined as an estimate (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human population (including sensitive subgroups) that is likely to be without appreciable deleterious effects during a lifetime.[504] It is intended to represent the level at or below which adverse (noncancer) health effects would not be anticipated, barring other concurrent exposures and/or compromising factors. Thus, exposure at the RfD for the associated exposure frequency/exposure duration combination (acute/subacute, subchronic, or chronic) should not produce adverse health effects. The converse, however, is not necessarily true; that is, exposure exceeding the RfD does not mean that adverse health effects will occur, although they may. It is true to say that the higher the exposure above the RfD, the greater the likelihood of adverse health effects occurring.
The critical effects listed in Tables 92 through 100 are the toxicity endpoints which provide the basis for each associated toxicity value listed. The most common endpoint listed is cholinesterase (ChE) inhibition. "Cholinesterases," as used here, refers to two enzymes found in animals, including humans: acetylcholinesterase (AChE) and butyrylcholinesterase. The inhibition of the critical enzyme acetylcholinesterase in the central and/or peripheral nervous systems provides direct evidence of potential adverse health effects. However, in the absence of other signs and symptoms, inhibition of nervous system acetylcholinesterase alone is not an adverse health effect. Likewise, inhibition of blood cholinesterases is not itself an adverse effect, although it may indicate a potential for adverse health effects in the nervous system. Red blood cells (RBCs) contain only acetylcholinesterase, while blood plasma contains both butyrylcholinesterase and acetylcholinesterase in varying ratios depending upon species. The EPAs Office of Pesticide Programs (OPP) asserts the following regarding cholinesterase inhibition data: 1) the best data are measurements of acetylcholinesterase inhibition in specific regions of the brain; 2) whole-brain measurements, while potentially useful, may not be sensitive to critical changes in discrete regions; 3) RBC measurements are preferred over plasma measurements; and 4) plasma measurements, despite shortcomings, can be useful in addition to other measurements. For the past several years OPP has required a neurotoxicity screening battery or separate studies for pesticide product registration which characterize the time course of inhibition in plasma, RBCs, and brain, including in specific brain regions, after acute and 90-day exposures. However, existing data sets usually lack the regional brain data.[505]
The EPA has identified only chronic oral values for DEET (Table 94). They have not identified any other relevant toxicological endpoints for the quantitative risk assessment of DEET.[506] The EPA has been evaluating DEET extensively over a period of years, and has concluded that DEET insect repellents will generally not cause unreasonable risks to humans when used according to label directions.[507] It has been used by millions of people since first investigated in the 1940s; today approximately 30% of the US population uses DEET annually, and few, if any, adverse health effects can be attributed to its use with any certainty.[508] There is some evidence that DEET may facilitate the absorption of other chemicals such as pyridostigmine bromide (PB)[509] however, typical risk assessment methodology, as used here, is not well suited to addressing this issue. Additional details regarding DEET toxicity are available in the RAND Literature Review.[510]
The RfDs used to evaluate chlorpyrifos were obtained from, or derived based on, a preliminary risk assessment document issued by OPP on October 18, 1999.[511] These values are the most relevant available, since they are based on cholinesterase suppression in adults. OPP has recently identified a more conservative toxicity value known as a population adjusted dose (PAD) for chlorpyrifos incorporating a Food Quality Protection Act (FQPA) factor of 10 for protection of children, embryos, and fetuses.[512] The PAD was not applied in the HRA.
(2) Other Human Benchmarks
Other human benchmarks, supplementing the standard toxicity values, are provided in Tab J. These benchmarks are based solely on human studies, and provide another means to characterize the potential hazards associated with the pesticide active ingredient doses estimated in the exposure assessment, although they have not been verified by an outside agency for purposes of risk assessment. Investigators have completed the compilation of benchmarks for the 12 POPC active ingredients. The summary of benchmark values and associated information are presented in Table 104.
The benchmarks presented in Tab J vary widely in terms of reliability for risk assessment purposes. Some of the benchmarks are based on controlled studies with multiple dosing groups and several individuals at each dose level. The doses and timing are certain. In contrast, other benchmarks are based on uncontrolled exposure studies in one or only a few individuals, where doses and timing are estimated but uncertain. The endpoints observed across all benchmarks identified span a wide variety of signs and symptoms, including cholinesterase suppression, mild to moderate frank effects, such as reduction of tendon reflexes, and severe effects, such as coma and death. Investigators identified benchmarks as acute/subacute, subchronic, and chronic, consistent with foregoing portions of the HRA; however, there are many gaps in the literature.
Table 104. Other human benchmarksa
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
b. Carcinogenic Effects
Tables 101 through 103 present the slope factors and associated information for the evaluation of carcinogenic effects. Carcinogenic effects are assessed by application of slope factors (SFs). An SF relates a dose to a probability of excess cancer. In the present case, "excess" cancer refers to cancer hypothetically caused by the exposure to pesticide active ingredients during the Gulf War, as distinguished from cancers due to all other exposures over the course of a lifetime. It should be borne in mind that about 40% of Americans will develop cancer during their lifetimes.[573]
For carcinogens, the EPA has developed weight-of-evidence classifications which are reported along with SFs. The weight-of-evidence classification developed by the EPA reflects the likelihood that an agent is a human carcinogen based on available data. There are five groups into which constituents may be classified with regard to carcinogenic potential:
A - Human carcinogen.
B1 or B2 - Probable human carcinogen. B1 indicates that limited human data are available. B2 indicates sufficient evidence in animals and inadequate or no evidence in humans.
C - Possible human carcinogen.
D - Not classifiable as to human carcinogenicity.
E - Evidence of noncarcinogenicity for humans.
Occasionally the exact classification is not clear, and the EPA may list a group range, such as "B2/C," meaning that the weight of evidence is not sufficient to place the pesticide active ingredient in one specific group. Also, alternative classifications may be used. For example, OPP states that malathion has "suggestive evidence of carcinogenicity but not sufficient to assess human carcinogenic potential."[574]
c. Provisional Toxicity Values
The largest number of EPA-verified toxicity values for pesticide active ingredients are for oral exposure, followed by inhalation and dermal exposure. However, where there are no suitable published toxicity values for a given type of exposure, published values for another type may be used as is or modified for use, if appropriate, to generate "provisional" values. For example, the EPA has suggested that in some cases it is appropriate to modify an oral reference toxicity value to reflect dermal absorption.[575] All provisional values we used are conservative. The level of confidence in calculated risks based on provisional values is generally lower than with published verified values.
Oral toxicity values are usually presented as administered-dose values. The calculation of hazards and risks via the dermal route, however, requires the use of absorbed-dose values. Therefore, the administered-dose values must be converted to absorbed-dose values. To convert an administered-dose RfD to an absorbed-dose RfD, the administered-dose value must be multiplied by the appropriate oral absorption factor. To convert an administered-dose slope factor to an absorbed-dose slope factor, the administered-dose value must be divided by the appropriate oral absorption factor.
Table 105 presents the oral absorption factors used for the POPCs, taken from the sources listed. The absorption of a pesticide active ingredient through the gastrointestinal tract in either humans or test animals is dependent upon various factors, including the concentration of the pesticide active ingredient, diluent, and the nutritional status of the individual.
Table 105. Oral absorption factors[576]
|
|||||||||||||||||||||||||||||||||||||||
|
In many cases the provisional values listed are simply values published for other exposure conditions. For example, where appropriate, a published chronic value was used as a provisional subchronic value. Since a chronic value is the most conservative, its use as a subchronic or acute/subacute value enhances conservatism. Likewise, in some cases a published oral value was used for an inhalation value. The latter is conservative in the sense that the inhalation route is evaluated, but it may either underestimate or overestimate the true toxicity via inhalation.