Pesticide Exposure Risk Assessment: Hazards, Exposure, and Children's Susceptibility - Pro, Exams of Entomology

Download Pesticide Exposure Risk Assessment: Hazards, Exposure, and Children's Susceptibility - Pro and more Exams Entomology in PDF only on Docsity! ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 1 of 23 November 5, 2003 Lecturer: Allan Felsot, Department of Entomology, Food & Environmental Quality Lab, WSU- TC campus; afelsot@tricity.wsu.edu Issues in Pesticide Hazards—Children; Cancer; Endocrine Effects I. Overview of Risk Assessment A. A necessary first step to determining the likelihood (i.e., risk) that exposure to a pesticide might cause an adverse effect is to review the process of risk assessment. 1. Often terms like hazard, exposure, and risk are misunderstood and occasionally used interchangeably. a. However, each term has a specific meaning in the risk assessment process, which is applicable to pesticide regulation and eventually to permitted uses. 2. Furthermore, risk assessment and risk management are sometimes confused, perhaps because risk assessment has elements of both science (in the broadest sense of the process of hypothesis testing) and management (in the sense of policy, which is influenced by politics, economics, and sociological factors). B. Risk assessment as currently practiced by the EPA in its regulation of toxic substances (and recommended by the National Academy of Sciences) consists of four processes: 1. Hazard identification (or assessment and characterization) 2. Dose-response assessment 3. Exposure characterization 4. Risk characterization C. The first three processes in risk assessment are scientific in nature; i.e., experiments are conducted using sound principles of hypothesis testing to determine (characterize) toxicological hazard, the relationship between dosage and response (adverse effects), and degree of exposure in the environment. Risk characterization, on the other hand, is partly scientific and partly management. D. Hazard Assessment: determination of the range of possible biochemical and physiological responses to a toxicant. The process necessarily involves administering low and high doses to achieve a full range of effects. 1. Hazard assessment studies are the most common type of experiments found in the toxicological literature. a. Most of the studies are motivated by the objective of determining the mechanism of toxicity. 1. Thus, these studies are most often characterized by few doses, sometimes only one, administered to an animal. 2. Often to achieve an easily measured effect, the doses are given by injection, either intraperitoneally or subcutaneously. 3. Other variables in dosing include acute (single dosing), short term (perhaps during gestation only, or several weeks), subchronic (usually 13 weeks or 90 days), and chronic (life-time equivalent; the average lifespan of a rat is considered about 2 years). ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 2 of 23 b. Fewer studies are governed by the objective of discovering a dose causing no effect. 1. Ironically, this “negative” data, although less likely to be published by members of research-oriented institutions like universities, is the information most valuable to regulatory agencies for proper risk assessment. a. Industry, however, conducts and submits the research that includes the lowest dose not causing an effect (No Observable Adverse Effect Level, NOAEL) for the various hazards that are required to be tested prior to pesticide registration. 2. Similarly, few investigators ask the question directly, why did this substance cause no effect at this dose but an effect at a higher dose? a. Toxicokinetic studies are valuable in answering such a question. 2. Information about hazards of substances comes from three basic types of experiments: a. In Vitro 1. Tests are conducted on purified or gross tissue homogenates of enzymes and receptors; cell and tissue cultures; perfused organs a. Dosages are difficult to relate to whole body exposures unless toxicokinetics are well understood. b. Mutagenicity studies, or the potential for causing genetic changes, are often in vitro studies, but can be studied following a dose to a whole animal. 1. One example of a routine mutagenicity test is the Ames Test that examines the number of reversions of a histidine-minus bacterial strain to a histidine-plus character in the presence of a toxicant. b. In Vivo 1. Live animals are administered doses by various routes, including dermal, inhalational, oral. 2. Subsequent measurements range from biochemical effects to behavioral effects. c. Epidemiological 1. Epidemiology, which evolved to study incidence of pathogenic diseases and causes of the disease, relies on Koch’s postulates. 2. Post WWII, the traditional role of epidemiology was stretched to include chemical substances, but in contrast to positive identification of a disease organism and traceability of exposure levels, chemical epidemiology suffers from inaccurate (and imprecise, or even non-existent) exposure information. 3. Hazard assessment studies examine the widest possible range of toxicological endpoints. The required tests and guidelines for conducting them are promulgated by the EPA and are considered validated procedures for conducting studies that are to be submitted in support of a petition for pesticide registration (see “OPPTS Test Guidelines Series 870, Health Effects” http://www.epa.gov/docs/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/ Series/). The tests include variations of the following hazards: a. Death 1. Single dose (acute exposures); estimation of LD50 ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 5 of 23 high doses, weight loss (defined as significant if 10% or more compared to control groups) is often the only toxicologically significant endpoint. F. Exposure Characterization 1. Estimate or direct measure of how much pesticide a person (or nontarget organism) contacts; a. Does not take into account toxicokinetics unless extrapolations are being made from one pathway of exposure to another; 1. For example, if an oral exposure toxicity study is used to determine hazards and potential dose causing toxicity from dermal exposure, than the oral dosage is multiplied by the absorption efficiency. 2. Implicit, therefore, is the assumption that all of an oral dose is absorbed into the blood and is distributed to the tissues. 2. The FQPA mandates that exposure of pesticides to consumers be aggregated from the diet (i.e., food), water, and residential (indoor or outdoor) use. 3. Food residues are directly measured. If not, than the EPA assumes the tolerance level. 4. Water residues are almost always generated from computer simulation models, even though the USGS has accumulated a database of residues monitored in major watershed basins throughout the U.S. 5. Residential exposures are either based on direct measurements, or extrapolations from a database of occupational exposure studies known as PHED (Pesticide Handler and Exposure Database). G. Risk Characterization 1. Risk characterization is partly scientific and partly management (policy). a. For example, from dose-response assessments, one can observe directly a dosage that causes no adverse effect in the test animal. This dosage is known as the NOAEL (No Observable Adverse Effects Level). 1. A level of exposure either estimated or directly measured in the environment can be compared to the NOAEL, forming a ratio that describes how much under or over the NOAEL of a contaminant or drug a person is being exposed to. This ratio is called the margin of exposure (MOE). Margin of Exposure (MOE) = NOAEL (mg/kg/day) Exposure (mg/kg/day) 2. For exposure to pesticides, the risk or likelihood of an adverse effect in the environment is characterized after deciding what an acceptable MOE would be. For consumer exposure, the EPA considers MOEs greater than 100 to be of no concern (i.e., the Level of Concern or LOC), meaning a reasonable certainty of no harm. 3. Another way to estimate risk is to compare exposure to the Reference Dose (RfD). The RfD is the NOAEL adjusted by a 100-fold safety (uncertainty) factor. Any exposure less than 100% of the RfD is below EPA levels of concern (LOC). Reference Dose (RfD) = NOAEL 100 ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 6 of 23 EPA Risk Estimate = Exposure (mg/kg/day) RfD x 100 ; (which must be less than 100% of RfD.) a. EPA uses the MOE approach to risk characterization when dealing with worker exposure and residential exposure. The MOE approach is also used when characterizing risk from aggregate and cumulative exposure. b. The “percentage of RfD” approach is used when only examining the risk associated with dietary exposure. c. Determining an appropriate safety factor is subjective, i.e., it is management. There are no scientific principles that would dictate the use of one safety factor over another; however, some have argued that safety (i.e., uncertainty) factors can be derived by examining the available data on the range of endpoint response among animals (for example, comparing the range of neonate LD50’s to the adult LD50’s; examining the differences in toxicokinetics and/or toxicodynamics among animals or between animals and humans) (Dourson, M. L., S. P. Felter, and D. Robinson. 1996. Evolution of science-based uncertainty factors in non-cancer risk assessment. Regulatory Toxicology & Pharmacology 24:108-120) 1. Nevertheless the rationale for using a standard 100-fold safety factor in translating the NOAEL into the risk parameter known as the Reference Dose (RfD) is a 10-fold factor for translation of data from rodents to humans (in case humans are more sensitive than rodents) and another 10-fold factor for response variability in the human population (in case children and seniors are more sensitive than middle aged men.) 4. Risk characterization when children are deemed more sensitive at a given dose than adults; a. Note that the FQPA specifically mandates EPA to make a determination of whether infants and children are more sensitive to a given dose of a pesticide than adults. b. If the findings are affirmative, than EPA uses up to a 10-fold additional safety factor to estimate the RfD. At this point, the RfD is transformed into the Population Adjusted Dose (or PAD). 1. Similarly, if the MOE method is used to characterize risk, then the acceptable MOE will be 1000 rather than 100. Population Adjusted Dose (PAD) = RfD 10 , which is substituted into the EPA Risk Estimate equation above. 2. Note that the acceptable risk is still considered below 100% of the PAD. II. Why the Focus on Children? A. In a nutshell, kids are not little adults. That is to say, they have major morphological, physiological, and behavioral differences that could expose them to more pesticide. 1. But we are also concerned that the mode of toxicity or the rate of metabolism might be different, creating increased hazard in comparison to adults. ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 7 of 23 2. You will recall from the previous lecture, that the National Academy of Sciences Report, Pesticides in the Diets of Infants and Children, gave rationale to the provisions put into place by passage of the FQPA. B. Below is an excerpt from an essay that I wrote that concisely explains some of the differences between children and adults [Felsot, A. S. 1999. Pesticides, children, and the FQPA: Where are we after fifty years of exposure? ? Agrichemical & Environmental News (January) 153:6-9.]. (Archived issues available at http://aenews.wsu.edu) “Children Are Not Little Adults The FQPA states that EPA will “ensure that there is a reasonable certainty that no harm will result to infants and children from aggregate exposure” to pesticide residues. Typically, what is a reasonable certainty is left to the regulatory agency, and the question quickly turns to risk management rather than to risk assessment. The latter is where the science occurs while the former is based primarily on considerations of politics, economics, values perhaps informed by scientific principles. With regard to any type of chemical exposure, however, sound scientific principles do support special consideration of infants and children. Pediatricians have long known that the physiology of infants, children, and adults are different, and such differences could influence the therapeutic doses of medicines as well as the doses that are hazardous. Notable differences that may influence toxicity of drugs and other chemicals include the following.” • Infants and children have a greater surface area to body mass ratio than do adults. • Brain size in infants and children is proportionally greater relative to body mass than in adults. • The fetal brain is rapidly developing, laying down new nerve connections especially in the third trimester of pregnancy and than continuing rapidly during the first year after birth. The network of connections between nerve cells is believed to be dependent on chemical neurotransmitters like acetylcholine. • The blood-brain barrier, a membrane-like tissue rich in blood vessels surrounding the brain, is less developed than in adults and not as impermeable to certain chemicals. Also, cerebral blood flow is greater in children than in adults. • The ventilation (breathing) rate of infants is significantly greater than in adults, resulting in greater inspired air exposure per unit time. • Filtration of the blood through the kidneys (known as renal clearance) is slower in infants than in adults. • Enzymes known to detoxify chemicals may be at lower levels in infants and children than in adults.” “Are Children More Sensitive to Chemicals Than Adults? The differences in physiology notwithstanding, generalizations about chemicals being more hazardous to children are not possible. Some chemicals may cause greater toxicity in children than in adults, but the reverse situation could also be true. For example, some chemicals are activated to toxic products by a special group of enzymes called mixed function oxidases (MFOs). If levels of MFOs were low, a toxicant may not become activated sufficiently to cause an adverse reaction. On the other hand, certain enzymes, including the MFOs, also break down ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 10 of 23 computer servers or nodes at key places in the world. Thus, your individual computer may actually be communicating with a particular node that then sends the messages to another node or directly to another computer. The receiving node or computer can give feedback to the sending node or computer, establishing a two-way interactive system of communication. Meanwhile, the individual computer initiating the message, has within itself its own parts, like the CPU (central processing unit), ROM (read-only memory), the hard disk, and the video screen that involve two-way communication among themselves. Think of the endocrine system as one node of a body internet that also includes the nervous system and the immune system as the other nodes. Each of these nodes communicate within their own system and with each other. While computers within themselves and across the Internet communicate via a combination of electrical signals and telecommunication microwaves beamed to satellites and back, the body’s internet communicates information across the nodes via a system of chemical messengers. The chemical messengers of the endocrine system are called hormones. Hormones are produced by ductless glands and released (secreted) directly into the blood stream. The most well-known hormones are sex steroids like estrogen, produced by the ovaries, and testosterone, produced by the testes. Estrogen and testosterone are also produced in the adrenal gland associated with the kidney of both sexes. Testosterone can be changed into estrogen by an enzyme called aromatase that is prevalent in brain cells. Thus, males also have low levels of estrogen as well as testosterone in the blood. Some of the other well-known hormones include those produced by the thyroid (thyroxin) and pancreas (insulin). The speed with which hormones can work, and an indication of their ability to communicate with the brain, is illustrated by how fast one flinches when they see an object coming straight toward them. This behavior is mediated by the hormone adrenaline, produced in the adrenal glands and in the brain. In the nervous system, the messengers are called neurotransmitters and are released at the endings or junctions between nerves or between nerves and muscle or glands. These junctions are actually very tiny physical spaces into which chemical messengers are released. For example, acetylcholine is a chemical released at nerve endings that permit an electrical nervous impulse traveling across the fibers of a nerve to be transmitted to an adjacent nerve, where a new electrical signal is propagated. An intimate connection between the central nervous system (i.e., the brain and spinal cord) is made through the communication of neurotransmitters with the hypothalamus and pituitary, two endocrine glands in the brain. When stimulated, these glands release hormones that circulate throughout the body to affect other organs, including the sex glands. The thymus gland, lying near the heart, is the master controller of the immune system, regulating the production of the myriad immune cells and antibodies. Certain immune cells produce hormones called cytokines that can interact with the brain. In response, the brain may produce hormones that affect other glands and organs, including the thymus.” “Translating the Message From the blood, hormones interact with cells by binding to special proteins called receptors. Many organs and glands contain receptors for one or more hormones. Receptors are located in membranes on the outer cell surfaces or on the nucleus. The nucleus of the cell contains all the genetic information, which is stored in the DNA, the information containing biochemical polymers making up the genes. The chemical messenger actually binds with the receptor like a lock and key. When enough receptors are bound by the messenger, the lock is ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 11 of 23 opened and a signal is transferred to the DNA. The DNA is “woken” up, setting in motion a chain of events that causes the genes to produce proteins necessary for proper functioning of the individual cells, tissues, and organs. Translation of the message by target cells result in physiological reactions ultimately responsible for stimulating, regulating, and maintaining proper metabolism, development, growth, reproduction, and behavior. The hormones stimulate physiological responses at incredibly minuscule concentrations. For example, estrogen can stimulate growth of cells at levels equivalent to parts per trillion. Furthermore, the timing of the messages, is crucial to normal development, especially in the fetus. Thus, the right amount of hormone must be present at the right time for a male or female to develop normally. Studies with mice have shown that higher than normal amounts of estrogen at the wrong time during pregnancy can cause genetically male rats to behave more like females, and in some cases to develop female-like genitals. Hormone concentration and timing is crucial not only to sexual development, but also to normal development of the brain. A malfunctioning endocrine system during fetal or infant development could potentially alter the proper functioning of the immune system in later adult life. In short, the endocrine system and its interactions with the brain and immune system are exquisitely balanced and timed.” “A Promiscuous Flaw The various hormones function like keys for specific receptors by virtue of their three dimensional molecular structure. Unfortunately, the receptors can also be activated by chemicals whose molecular structure mimics the natural hormones. Chemicals with this ability are called endocrine disrupters, and they can act in a wide diversity of ways as indicated by EPA’s working definition--endocrine disrupters “interfere with the synthesis, secretion, transport, binding, action, or elimination of natural hormones in the body that are responsible for the maintenance of homeostasis [normal cell metabolism], reproduction, development, and/or behavior.” The list of chemicals with endocrine disrupting potential is growing but contains the usual suspects: various pesticides, persistent chlorinated organics like PCBs and dioxins, plasticizers, surfactants, and heavy metals. But the list also includes natural biochemicals: mycotoxins produced by fungi and chemicals produced by fruit and vegetables (called phytoestrogens). EPA’s definition of an endocrine disrupter covers a wide array of possible effects. A discussion of the concerns can be found in the EPA special report, “Environmental Endocrine Disruption: An Effects Assessment and Analysis.” Most of the media attention has been focused on chemicals acting like estrogen, but this is only one mode of action. More important are the potential endpoints or diseases in humans: breast cancer and endometriosis in women, testicular and prostate cancers in men, abnormal sexual development, reduced male fertility, alteration in pituitary and thyroid gland functions, immune suppression, and neurobehavioral effects. While many of these adverse effects in humans are only hypothesized as being associated with endocrine disrupters, some scientists are arguing that enough evidence as accumulated to conclude a cause and effect relationship between endocrine disrupters and disease in wildlife. Adverse effects include abnormal thyroid function and development in fish and birds; decreased fertility in shellfish, fish, birds, and mammals; decreased hatching success in fish, birds, and reptiles; demasculinization and feminization of fish, birds, reptiles, and mammals; defeminization and masculinization of gastropods, fish, and birds; decreased offspring survival; and alteration of immune and behavioral function in birds and mammals.” ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 12 of 23 “Dose Still Makes the Poison In past essays I have written about the use of extremely high dosing required by the EPA for tests of carcinogenic potential of pesticides, and I have questioned the validity of the results for assessing the effects of exposure to the very tiny environmental concentrations. I am ready to make similar conclusions about the applicability of endocrine disrupter screening to real world exposures. The current laboratory testing procedures for endocrine disrupters do show that certain chemicals can activate the estrogen receptor or block activation of the testosterone receptor. However, the doses required to show these effects in a test-tube type experiment are thousands to millions of fold greater than for the natural hormones. Such observations suggest that environmental concentrations of the “endocrine disrupters” may be irrelevant to producing a biological effect. Endocrine disrupting potential of pesticides like DDT have been tested by directly feeding it to rats. Adverse effects on sexual development have been reported, but the dosing rate was millions of times greater than what humans are normally exposed to in food. Thus, with regard to testing methodology, endocrine disrupters seem to be in the same boat as carcinogens. Research reported by Tulane University in 1996 suggested that perhaps very low doses of endocrine disrupters were physiologically important. The researchers purportedly showed that two or more chlorinated hydrocarbon pesticides administered together could have a synergistic endocrine disruptive action at doses a thousand-fold lower than the doses causing the same activity when alone. When dieldrin and endosulfan were dosed together at levels equivalent to parts per billion, an estrogenic effect was observed; when given alone, parts per million were required to produce the same effect. However, no other laboratory was able to duplicate the results of the synergism experiment, and finally, the researchers at Tulane themselves admitted they could not repeat their own observations. Nevertheless, Tulane’s retraction has not quieted the storm. Whether or not there are synergistic interactions among low doses, scientists still have shown that doses of two or more chemicals can be additive, albeit the concentrations necessary for an effect are extremely high. Perhaps stronger cases for widespread problems linked to endocrine disrupters will be made when the dosing becomes more realistic and thresholds for effects are determined. Meanwhile, the FQPA requires EPA to certify screening tests for endocrine disrupters. Eventually all pesticides will be subjected to the approved battery of tests.” B. Note that during 1999, the National Academy of Sciences, through its research arm, the National Research Council (NRC), issued a report, Hormonally Active Agents in the Environment. 1. The NRC concluded that some biochemical/physiological effects could not definitely be concluded to be due to hormone disruption. Thus, the committee decided it was better, and perhaps more accurate to exchange the term endocrine disrupters for hormonally active agents (HAAs). a. Thus, far, most HAAs, are first earmarked as hormonally active in in vitro tests. However, in vivo tests must be conducted to understand whether the compound is also active when toxicokinetic factors are at work. 2. One major conclusion from the NRC report was that evidence was too weak to conclude that exogenous HAAs had affected human health. The evidence was much ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 15 of 23 a. Furthermore, epidemiological studies at best can make associations between exposure to an agent and an outcome, but this is quite distinct from concluding an agent caused an outcome 1. Epidemiological studies rely heavily on regression analysis 3. Epidemiology grew out of the need to control infectious diseases, which have definable causes; furthermore, infectious diseases, which are essentially microbiological problems, can be directly tested using Koch’s postulates, stated as follows: a. The infectious agent (microbe) must be present in every case of a disease; b. The microbe must be isolated from the disease and grown in pure culture; c. The specific disease must be reproduced when a pure culture is inoculated into a healthy susceptible host; d. The microbe must be recoverable again from the newly infected host. 4. Unfortunately, for most chemical exposures, unless they are acute (as opposed to chronic), Koch’s postulates are not applicable. 5. Furthermore, in many chemical epidemiological studies that do not involve workers at a specific industry (or manufacturing site), exposure records are poor to nonexistent. Often exposure is deduced from interviews of “what was used” or “next- of-kin” interviews. F. The attempt to associate chemicals as etiological agents with specific diseases has had some successes, but the widespread use of epidemiology to make solid cause and effect relationships rests on tenuous ground. 1. An example of a success is the association of scrotal cancer with the profession of chimney sweep. 2. The evidence gained from human studies is strengthened by consistency among several studies a. Conflicting results among well-done, large epidemiological studies raise serious doubts about apparent associations b. One limiting factor is dose quantification; 1. Some investigators use qualitative exposure estimates that raise problems of misclassification of exposure a. low, high, medium b. use of a chemical for X number of years G. Measures of Association (information from Draper, 1994, ACS Advance in Chemistry Series 241, Environmental Epidemiology) 1. Relative risk a. A measure of how many times greater the risk for one population is compared to another population; 1. Commonly used with cohort type studies where large numbers of people who have been considered for their exposure at some specified base-line time and are then subsequently observed for the development of disease. a. If disease is related to exposure, the frequency of exposure should be greater among the diseased than among the non-diseased group. 2. The incidence among those exposed to a risk factor (Ie), divided by the incidence among those not exposed (Io) ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 16 of 23 a. RR = Ie/Io (or, incidence of disease in exposed group divided by incidence of disease in unexposed group). b. RR is a measure of the strength of an association. 1. The greater the RR value, the more likely that the risk factor is important in causation c. Generally, we tend to doubt relative risks of less than 1.5. 2. Attributable risk a. Measure of the impact of a risk factor b. Defined as the difference between the incidence in the exposed and the unexposed--or, the excess incidence among the exposed after removing the expected background incidence 1. AR = Ie - Io 2. AR is the portion of the incidence among those exposed that can be attributed to that exposure Relative vs. Attributable Risk Factor Lung Cancer Coronary Heart Disease Heavy smokers (incidence) 166/100,000 599/100,000 Non smokers (incidence) 7/100,000 422/100,000 Relative risk (ratio) 23.7 1.4 Attributable risk (cases/100,000) 159/100,000 177/100,000 3. Odds Ratio a. A measure of relative risk for case-control studies. 1. Utilize samples of diseased and non-diseased persons to determine the frequency of the exposure of interest, rather than to evaluate exposure in a disease-free population and to await subsequent disease expression as would be done in a cohort study. 2. Odds Ratio (OR) = ~ (incidence of exposure in diseased group/incidence of exposure in non-diseased group) a. Thus the OR represents the frequency or gradient of exposure in the diseased group relative to the frequency of exposure in the control or non- diseased group. 1. For example, if you hypothesize that compound X causes Non- Hodgkin’s Lymphoma (NHL, a type of cancer attacking the lymph glands) among a certain occupation using compound X, than if you compared the frequency of compound X use among NHL sufferers, you would predict a higher incidence of compound X use than in nonusers without NHL. H. Measures of Uncertainty 1. For measures of OR and RR, a 95% confidence interval (CI) about the average risk ratio is usually given in a report. The 95% CI represents the probability (i.e., less than 5% error) that the true risk has been captured in the stated interval if it was repeatedly sampled. For example, if an experiment was conducted 100 times, and the error rate less than 5%, you would have captured the true population risk in 95 of those experiments. The CI would differ with each independent experiment, but the ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 17 of 23 probability of capturing the true mean would still be set at 95% (or at whatever probability the experimenter determines is appropriate). If the lower end of the 95% CI goes below a relative risk or odds ratio of 1, it is appropriate to conclude that no conclusion can be reached with regard to the true population mean being greater than 1 (i.e., a risk greater than 1). I. Cases where epidemiological studies work 1. When a sudden chemical exposure in which an illness is produced within a very short time occurs (for ex., the chemical plant explosion in Bhopal, India, which spewed large amounts of methyl isothiocyanate (MITC) 2. When under typical conditions and in very short periods of time, a researcher can correlate a change in a physiological function to the change in an environmental pollutant (for ex., children with asthma in areas where air pollution is caused by ozone) 3. When an association has been made between long-term exposures and long-term functional effects (lead and children; can measure lead deposited in teeth) 4. When occupational studies have linked relatively high levels of chemical exposure to the incidences of cancer in workers J. Unfortunately, unless a specific occupational exposure has occurred, or a specific accident has occurred, exposures to environmental contaminants are not amenable to good epidemiological assessments essentially because they lack the quality of the situations described above. 1. Lack of quantitative dose-response relationship at environmentally relevant doses; 2. Lack of proper exposure assessment; 3. Lack of specificity (disease could be caused by a lot of factors); 4. Confounding factors not perceived. K. A 1997 study by Daniels et al. (Environmental Health Perspectives 105:1068; “Pesticides and Childhood Cancers”) critically analyzed epidemiological studies published between 1970 and 1996 that involved associations with pesticide exposure and childhood cancer (n=31). 1. Any associations were moderate (meaning they were less than an Odds Ratio of 2.0), but the strongest associations were seen between brain cancer and leukemia and frequent occupational exposure of parent to pesticides or home pesticide use. 2. However, examining the data tables from brain cancer and leukemia, I noted the following: a. For childhood brain cancer, 40 correlations were shown (thus, some studies had more than one correlation in them). Of these 40, only 7 had lower 95% confidence intervals >1.0. b. For childhood leukemia, 25 correlations were shown. Only 4 had a lower 95% confidence interval greater than 1.0. L. One more aspect to consider is biological plausibility. 1. Many chemical epidemiological papers fail to consider the myriad of rodent chronic exposure studies that have been submitted to the EPA for consideration of pesticide risk assessment. a. These studies often have NOAELs for excess tumor occurrence. b. Furthermore, reference doses are estimated and exposure is not supposed to exceed this dose if the pesticide is to be registered. ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 20 of 23 occurred—tearing, salivation, and tremors—indicating that MTDs are still quite toxic. Many studies administered doses by subcutaneous (under the skin) or intraperitoneal (into the abdomen) injection. Such unconventional exposure routes expose an organism to a very large dose all at once, bypassing the protective layer of the skin or the much slower absorption into the bloodstream from the intestine. Such drastic doses and methods of exposure are part of what I like to call mechanistic toxicology studies. Their objectives are to characterize physiological responses and their biochemical basis. Occasionally, doses are used that elicit no response in an attempt to track subtle changes in biochemistry. The objective of regulatory toxicological studies, on the other hand, is to find the NOEL for the most sensitive biochemical or physiological response. The DAS studies submitted for EPA review used doses spanning from the NOEL through those known to produce an effect. Chlorpyrifos doses were given directly to a mother rat during pregnancy and for 11 days after birth to expose both the fetuses and the suckling newborns (neonates). The parental NOEL was always lower than the NOEL for the neonates (Table 1). In other words, the neonates were less susceptible than the adults for the most sensitive endpoint examined. The endpoints ranged from enzyme inhibition to brain histopathology and functional behavior, all effects that would be predictive of adverse neurodevelopment.” “Why Dose Matters In nearly every case that EPA used to support its conclusion that infants may be more susceptible than adults, doses were incredibly high relative to real world exposures. For example, in one cited study 1-4 day old rats were injected subcutaneously with 1 mg/kg/day of chlorpyrifos (18), the MTD previously observed not to cause outward signs of anticholinesterase toxicity. Nevertheless, brain AChE was significantly inhibited. More importantly, the exposure was nearly 800 times greater than aggregate (dietary and residential) exposure both modeled as well as measured at the 99.5th and 100th percentile, respectively (6). Recent research compared neurochemical effects of chlorpyrifos and methyl parathion in neonatal and adult rats, and it sheds some light on the discrepancy in observations between the higher dose mechanistic studies and the lower dose regulatory studies (10). Each insecticide was subcutaneously injected into neonatal and adult rats for 7 or 14 days in a row with doses equivalent to about 20% of their MTDs. Brain cholinesterase and receptor binding inhibition were similar in neonatal and adult rats exposed to chlorpyrifos, but reversed to control levels more quickly in neonates. Neonates were always significantly more sensitive than adults to the effects of methyl parathion exposure. Thus, neonatal rats respond differently to different OPs. Neonatal rats are more susceptible than adults to the lethal effects of acute high doses of chlorpyrifos, but are less sensitive to the subacute intermittent doses and as equally sensitive when exposed to subacute daily doses. In essence, the level of dose determines the differential sensitivity for chlorpyrifos but not for methyl parathion.” “A Monkey Wrench in the Works?… Curiously, EPA gave a cursory nod to an intriguing new area of research regarding possible effects of chlorpyrifos on neuronal cell replication and growth. Exposure of neonatal rats to subcutaneous injections of 1 or 2 mg/kg chlorpyrifos resulted in an inhibition of DNA synthesis and abnormal functioning of one component of the cell cycle control system called adenylyl cyclase (18, 19). However, significant brain cholinesterase inhibition occurred ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 21 of 23 immediately after dosing, so it is doubtful whether the effects were truly more sensitive endpoints than plasma cholinesterase (ChE) inhibition (10). On the other hand, a hypothesis currently in vogue is that chlorpyrifos may affect neurodevelopment through inhibition of neuronal cell (neurite) branching and rate of growth (16). A recently released study used a special nerve cell culture (PC12) and showed that chlorpyrifos and its nontoxic metabolite TCP can reduce overall neurite growth without significant inhibition of ChE activity (4). The research suggests that chlorpyrifos and other anticholinesterase compounds react at a molecular site different than the one responsible for inhibiting the enzyme.” “…Or a No-Brainer? One of the problems with cell culture studies is the difficulty of relating the concentration of a toxicant in the dish to the dose given to a rat. The PC12 cell study did show a definitive NOEL for neurite response to chlorpyrifos and TCP. The authors claimed that the concentration adversely affecting neurite growth was similar to brain levels of TCP reported in a study where pregnant rats were fed chlorpyrifos for four days at a rate of 3 mg/kg/day (6A). This oral dose is nearly 2500 times greater than the highest aggregate exposure to children (6). The reported PC12 cell culture study may not have realistically reflected the concentration of chlorpyrifos or TCP that would be in the neonatal rat brain following exposure to a dose equivalent to the developmental neurotoxicity NOEL (see Table 1). TCP was not detected in blood from 5 day-old lactating neonatal rats whose mothers were exposed to 1 mg/kg/day chlorpyrifos during pregnancy and for 10 days after birth. Exposure of a neonate rat during lactation is most relevant to human fetal development because the newborn rat brain is equivalent to the developmental stage of a human brain during the third trimester of pregnancy (18).” “A “Weight” and See Attitude EPA claims that they will assess risk and make registration decisions using a “weight of the evidence” approach. While concluding that DAS’ data indicated no increased sensitivity of infants relative to adults, the agency chose to delve into the published scientific literature and find the “weight” it needed. Yet, most of the studies it did cite actually showed that neonatal rats were only more susceptible to lethal acute exposures but less susceptible at nonlethal intermittent or daily exposures. Three recently published papers have examined the same literature among other pieces of evidence and have thrown their “weight” behind removal of an extra FQPA safety factor (2, 6, 14). With transparency seemingly ruling the EPA lately, anyone can read the chlorpyrifos risk assessment documents (http://www.epa.gov/oppsrrd1/op/status.htm) and make up their own mind. EPA has invited submission of comments for its consideration as it prepares the final re- registration decision. I’m waiting to see if the Tale of Two Sciences will continue.” Table 1. Summary of EPA’s Interpretation of Results from Dow AgroSciences Tests to Determine Sensitivity of Infants Test Doses (mg/kg/day) Duration of Exposure Parental NOEL Offspring NOEL Developmental Toxicity 0, 0.1, 3, 15 Gestation Days (GD) 6-15 0.1 15 ENTOM 558 Pesticide Topics Fall 2003 ENTOM 558 Hazard 11-5-03.doc Page 22 of 23 Developmental Neurotoxicity 0, 0.3, 1, 5 GD 6 to Lactation Day 11 >0.3 1 Reproductive Toxicity 0, 0.1, 1, 5 2 Generations 12 weeks 0.1 1 References 1. Adams, J. 1999. The child’s brain: prenatal development. http://www.brain.com/about/article.cfm?id=2230&cat_id+61 2. Albers, J. W. et al. 1999. Analysis of chlorpyrifos exposure and human health: expert panel report. J. Toxicol. Environ. Health, Part B 2:101-124 3. Brimijoin, S., and C. Koenigsberger. 1999. Cholinesterases in neural development: new findings and toxicologic implications. Environ. Health Perspectives 107 (Supplement 1):59- 64. 4. Das, K. P., and S. Barone, Jr. 1999. Neuronal differentiation in PC12 cells is inhibited by chlorpyrifos and its metabolites: is acetylcholinesterase inhibition the site of action? Toxicol. Appl. Pharmacol. 160:217-230. 5. Felsot, A. S. 1999. Pesticides, children, and the FQPA: where are we after fifty years of exposure? Agrichemical & Environmental News no. 153 (January):6-9. 6. Gibson, J. E., Chen, W. L., and R. K. D. Peterson. 1999. How to determine if an additional 10X safety factor is needed for chemicals: a case study with chlorpyrifos. Toxicological Sciences 48:117-122. 6A. Hunter, D. L., T. L. Lassiter, and S. Padilla. 1999. Gestational exposure to chlorpyrifos: comparative distribution of trichloropyridinol in the fetus and dam. Toxicol. Appl. Pharmacology 158(1):16-23. 7. Kang, B. C., C. W. Wu, and J. Johnson. 1992. Characteristics and diagnoses of cockroach- sensitive bronchial asthma. Annals of Allergy 68:237-244. 8. Lan, J. L., D. T. Lee, C. H. Wu, C. P. Chang, and C. L. Yeh. 1988. Cockroach hypersensitivity: preliminary study of allergic cockroach asthma in Taiwan. J. Allergy Clinical Immunology 82:736-740. 9. Lauder, J. M., and Schambra, U. B. Morphogenetic roles of acetylcholine. Environ. Health Perspectives 107 (Supplement 1):65-69. 10. Liu, J., K. Olivier, and C. N. Pope. 1999. Comparative neurochemical effects of repeated methyl parathion or chlorpyrifos exposures in neonatal and adult rats. Toxicol. Appl. Pharmacol. 158:186-196. 11. Moser, V. C., S. M. Chanda, S. R. Mortensen, and S. Padilla. 1998. Age- and gender-related differences in sensitivity to chlorpyrifos in the rat reflect developmental profiles of esterase activities. Toxicol. Sciences 46:211-222. 12. Moser, V. C., and S. Padilla. 1998. Age- and gender-related differences in the time course of behavioral and biochemical effects produced by oral chlorpyrifos in rats. Toxicol. Appl. Pharmacol. 149:107-119. 13. Potera, C. 1997. Working the bugs out of asthma. Environ. Health Perspectives 105:1-4. 14. Schardein, J. L., and A. R. Scialli. 1999. The legislation of toxicologic safety factors: the Food Quality Protection Act with chlorpyrifos as a test case. Reproductive Toxicology 13:1-14. 15. Scheibel, A. B. 1997. Embryological development of the human brain. The Brain Lab @ http://www.newhorizons.org/blab_scheibel.html