![[logo] US EPA](../gif/logo_epaseal.gif){kind=logo}
A Consultation on Protocol Design to Assess Acute Neurotoxicity Following Oral Administration of Pesticides
presented by:
Dr. Kathleen Raffaele, Dr. Virginia Dobozy, and Dr. William Sette
Health Effects Division, Office of Pesticide Programs,
United States Environmental Protection Agency, Washington, D.C.
to the:
FIFRA Scientific Advisory Panel,
Arlington, Virginia
July 21, 1999
Background
In 1991 the Agency finalized a test guideline for screening chemicals for neurotoxicity. This Guideline was recently revised (OPPTS 870.6200, Neurotoxicity Screening Battery), and includes protocols for both acute and subchronic neurotoxicity screening studies. The acute neurotoxicity protocol is the only guideline protocol specifically designed to assess toxicity following a single exposure and to find a no-observable-adverse effect level (NOAEL) or minimum effect level. Hence, results from studies conducted under this protocol are often used by EPA's Office of Pesticide Programs (OPP) to select the endpoint to be used in acute dietary risk assessment for the general population. In acute neurotoxicity studies conducted under this guideline, an animal is given a single dose of test substance, and a variety of behavioral parameters are assessed at several time points prior to and following administration (pre-test, time of peak effect, 7 and 14 days after administration). Neuropathological effects are assessed at study termination. For specific chemicals, cholinesterase inhibition or other biochemical parameters may also be measured.
While the above guideline for conducting acute neurotoxicity studies does not specify a particular method of oral administration, such as gavage or dietary administration, it notes that for repeated exposures dietary administration will generally be acceptable. Also, other Agency acute and short-term oral toxicity test guidelines specify gavage as the preferred method for administering a test chemical. Two examples of test guidelines that specify gavage are the Acute Oral Toxicity guideline (870.1100) and the Acute Delayed Neurotoxicity Guideline (870.6100). The Developmental Toxicity Guideline (870.3100) also specifies gavage as the method of oral exposure, and requires justification for other routes or methods. Studies conducted in accordance with the developmental toxicity guideline use approximately 10 days of exposure in the rat, and the results are sometimes used for determining acute dietary endpoints since it is assumed that some developmental toxicity endpoints may result from single exposures. It is noteworthy that to date essentially all of the acute neurotoxicity studies submitted in support of pesticide registration have used gavage as the method of oral administration. Thus, gavage studies for assessing toxicity following acute or short-term exposures have long been the standard in regulatory toxicology.
Although the vast majority of the acute neurotoxicity studies submitted in support of pesticide registration have used gavage administration, several acute neurotoxicity studies have been very recently submitted to OPP in which the potential for acute neurotoxicity was evaluated following a one-time dietary administration over one hour. To summarize the general protocol during these studies, rats were first acclimatized to receiving their diet during two one-hour intervals of their daily dark cycle: one hour each at the start and finish of the dark cycle. Following acclimatization, test substance was administered as a single dose, mixed in the diet, during a one hour period at the end of the dark cycle. (Concentration of test substance in the diet was based on food consumption observed during the acclimatization period.) Food consumption was recorded, and administered dose was determined based on the amount of diet consumed by each animal. Behavioral parameters (and biochemical parameters, if appropriate), were evaluated at time of peak effect (determined by a pilot study) and at 15 days after treatment. In order to permit behavioral testing during time of peak effect, dietary periods were staggered within each test day, and test substance administration was also performed over several days (balanced by dose group). Although gavage administration is not specifically required under the current Agency acute neurotoxicity protocol several components that are required, such as day 7 testing and/or neuropathology assessments, were not conducted in these new, non-standard studies. While these studies were not conducted in compliance with the current Agency acute neurotoxicity guideline, this is not an issue for consultation with the Scientific Advisory Panel. The issues for consultation with the Panel are described below.
Issues for Consultation with the Scientific Advisory Panel
Issue #1 - Variability of Total Test Substance Intake Among Animals.
The first issue is variability of test substance intake in a dietary study. In a gavage study, the exact desired dose is directly administered to each individual animal, at a known time. In a dietary study, the test substance is mixed with the diet and the animal dosage varies with the actual amount of food consumed by each individual animal. In the submitted protocols, diet containing a specific concentration of test substance was offered to each animal during a one hour period. At the end of that hour, food was removed and consumption determined for each animal (including an estimate of spillage in at least one study). In order to achieve different dose levels for each group, the concentration in the diet was varied. In contrast to the exact dosing achieved in a gavage study, results from the submitted data indicate a highly variable intake within treatment groups. For example, in one dose group from one study, actual intake varied from 0 to 9 g of diet. Poor palatability could account for differences in amount consumed. With longer term studies, there may be initial palatability problems but animals accommodate to the taste of the test material in the diet with continued exposure.
Issue #2 - Variability of Test Substance Intake Patterns among Animals and Effects on Kinetic Parameters and Estimation of Time of Peak Effects.
In addition to variability of administered dose (Issue 1), the submitted protocols raise issues regarding the possibility of increased variability in measured response due to pharmacokinetic differences among animals (resulting from differences in consumption patterns within the one hour period). As noted above, test substance was offered to animals over a one hour period. Although no data were submitted regarding the pattern of consumption, it is likely that it varied among individual rats. For example, some animals may have consumed the diet immediately, some may have eaten slowly across the entire one-hour period, and some may have eaten quickly at the end of the one-hour period.
Since findings from these studies are based on evaluation at the 'time of peak effect', variability in the actual time at which test substance was consumed may lead to difficulty in detecting responses to test substance, especially for chemicals that are rapidly absorbed and have short duration of effects. For example, if effects peaked within 30 minutes of test substance consumption, time of peak effects would occur during the one hour dietary period if test substance was consumed immediately after it was offered, but 30 minutes after the end of the dietary period if test substance was consumed immediately prior to removal of diet. The magnitude of the peak effect might also vary depending on whether the test substance was all consumed within a short period or consumed slowly over one hour. This variability in time and magnitude of peak effect among animals would decrease the likelihood of detecting chemical-related effects, both due to flattening of the response curve across time (due either to flattening of individual response curves or to averaging responses with different peak times) and to the increase in variability in response among animals at any one particular time point.
Issue #3 - Effects of Restricted Feeding Schedules on Data Interpretation.
Changes in feeding schedules and light/dark cycles may cause changes in circadian rhythms, which may then cause behavioral changes in rats. In addition, normal feeding patterns vary among strains, and with age and sex of the animal. Because the submitted protocols involve changes in light/dark cycles and feeding schedules (and differences in dosing time in one protocol), behavior of the rats may vary from that in studies conducted using gavage administration. These differences may become especially important when chemicals with common mechanisms of toxicity are evaluated in cumulative risk assessments. In addition, if feeding schedules and light/dark cycles are altered during the study, behavioral assessments conducted prior to or following those alterations may not be comparable.
Issue #4 - Appropriateness of the Acute Dietary Exposure Model.
In conducting these new, non-traditional acute neurotoxicity studies, the registrants have asserted that administration of the test substance by dietary admixture is more analogous to expected human exposure than studies conducted using gavage administration. It is not clear whether that analogy depends more on the form (i.e. mixture with diet) or duration (i.e. a one-hour period) of test substance exposure. No data have been submitted to support the assertion that consumption of test substance in a dietary admixture over a one-hour period adequately simulates human dietary patterns. Human dietary patterns are likely to be highly variable. For example, human dietary patterns may include rapid (i.e., gavage-like) consumption of a glass of apple or orange juice or a breakfast bar, lunch eaten in 15 minutes, or dinner consumed over periods ranging up to several hours.
Questions for the FIFRA Scientific Advisory Panel
Issue #1 - Variability of Total Test Substance Intake Among Animals.
In a gavage study, the desired dose is administered directly to each individual animal. In contrast, in a dietary study the test substance is mixed with the diet, and the dose to an individual animal varies with the amount of food consumed.
a) How much variability in dose/consumption is acceptable within a single dose group (e.g. +10% of the mean)? If animals are included or excluded from analysis based on actual intake, does this bias the study results?
b) Is it possible to design an acute dietary study which ensures consumption of a uniform dose within each dose group? If so, please give examples of such designs.
Issue #2 - Variability of Test Substance Intake Patterns among Animals and Effects on Kinetic Parameters and Estimation of Time of Peak Effects.
In a gavage study, time and rate of administration are uniform for all animals. In an acute neurotoxicity study, effects are measured at the 'time of peak effects,' determined with reference to time of administration. In the submitted protocols, test substance is consumed in the diet over a one hour period. Variability in consumption patterns during that one hour period is unknown.
a) Does the Panel agree that mixing the test substance in the diet could change the pharmacokinetics of the substance? If so, how might this be assessed?
b) What impact might the temporal pattern of consumption (all at the beginning or the end of the interval, or steady eating throughout the interval) have on the magnitude and time course of the effects? Over what span of time of acute effects might this be a major factor (e.g. if peak effect occurred 6 hours after exposure, would the temporal pattern of consumption be less likely to influence results)?
c) How might these variations influence the estimation of the time of peak effect?
d) Could any such problems be resolved by alterations in protocol design? If so, please suggest appropriate modifications.
Issue #3 - Effects of Restricted Feeding Schedules on Data Interpretation
Normal feeding patterns vary with age and across strains of rats. In addition, altered feeding schedules are known to affect the rat's circadian rhythm, homeostatic indicators and behavior patterns. In order to achieve acute dietary intake (over one hour period), rats in the submitted protocols were acclimatized to restricted feeding schedules, consisting of two one-hour feeding periods each day (at the beginning and end of the dark cycle). In one study, the light/dark cycle was changed to 10/14 h light/dark, in order to accommodate the altered feeding schedule. Again, in one study, ad libitum feeding was resumed at varying periods following test substance administration.
a) Could these changes in feeding schedules and light/dark cycles differentially alter the study findings in the treated animals? What effect would changing these variables during the study period have on study results?
b) What influence would variations in normal feeding patterns (for example, due to strain, sex, or age) have on the results of studies using restricted feeding paradigms?
Issue #4 - Appropriateness of the Model
How well does a one hour dietary exposure period model human eating patterns?
Scientific Advisory Panel (SAP): July 1999Background Document for the Session